Tax Court Determines Values in Michael Jackson Estate Tax Case

ESTATE OF MICHAEL J. JACKSON, DECEASED, JOHN G. BRANCA, CO-EXECUTOR AND JOHN MCCLAIN, CO-EXECUTOR,
Petitioners
v.
COMMISSIONER OF INTERNAL REVENUE,
Respondent

UNITED STATES TAX COURT

Filed May 3, 2021

Avram Salkin, Steven R. Toscher, Robert Samuel Horwitz, Lacey E. Strachan, Howard L. Weitzman, Jeryll S. Cohen, Sharyn M. Fisk, Paul Gordon Hoffman, Edward M. Robbins Jr., and Loretta Siciliano, for petitioners.

Donna F. Herbert, Ray Malone Camp, Sebastian Voth, Jorden S. Musen, and Denise H. Larson, for respondent.

---

CONTENTS

FINDINGS OF FACT

OPINION

APPENDICES

---

MEMORANDUM FINDINGS OF FACT AND OPINION

HOLMES, Judge:

“The arts . . . proceed by the use of reason to the selection and adoption of what is appropriate, and to the avoidance and rejection of what is alien to themselves, contemplate the one class of objects with direct intent and by preference, and yet incidentally contemplate the other class also, and in order to avoid them.”¹

From the time he was a child Michael Jackson was famous; and there were times in his life, testified his executor, when he was the most famous person in the world. There were certainly years when he was the most well-known [*6] popular-music star, and even after his death there have been years when he was the world's highest-earning entertainer.

But there were also many years when he was more famous for his unusual behavior and not his unusual talent. And there were some years where his fame was turned infamous by serious accusations of the most noisome acts. We make no particular judgment about what Jackson did or is alleged to have done, but we must decide how what he did and is alleged to have done affected the value of what he left behind.

His Estate and the Commissioner agreed on the value of many of his assets, but continue to dispute the values of three intangible ones:

• Jackson's image and likeness;

• his interest in New Horizon Trust II, through which he held an interest in Sony/ATV Music Publishing, LLC; and

• his interest in New Horizon Trust III, which contained Mijac Music, a music-publishing catalog that owned the copyrights to compositions that Jackson wrote or cowrote, as well as compositions by other songwriters.

FINDINGS OF FACT

I. Early Life

Michael Jackson was born in August 1958. He was the eighth child in a family of modest means that lived in a small house in Gary, Indiana. Jackson [*7] began his career by singing with his brothers Tito, Jackie, Jermaine, and Marlon. He was not yet six years old.

Within a year Jackson became the lead vocalist for what became the Jackson 5. From late 1966 to 1968 the Jackson 5 played in nightclubs and music venues across the Midwest and Northeast. In 1968 the group auditioned in Detroit for Motown Records and its founder, Berry Gordy. They soon signed with Motown, and their first four singles on the label reached number one on the Billboard Hot 100 Chart — the first time that had ever happened.

The success of the Jackson 5 gave Jackson a springboard to launch his solo career. In April 1971 Jackson became the youngest individual ever to appear on the cover of Rolling Stone. In 1972 he released his first two solo albums, Got To Be There and Ben. The eponymous song “Ben” was the first of Jackson's solo songs to become a number-one best seller. It is proof of his talent or the oddity of the era's popular culture that it appears to be a song about the love of a boy for his rat.

In 1975 the growing popularity of Jackson and the Jackson 5 enabled the family to move from Motown to Epic Records, a subsidiary of CBS Records, and to change their name to “The Jacksons.” The Jacksons released albums and performed together until the mid-1980s. During this time Jackson's talents as a [*8] songwriter and composer emerged, and he wrote or cowrote 20 of the 84 compositions that The Jacksons recorded.

II. The Rise

Jackson had long had a difficult relationship — one he would later claim was abusive — with his father. His performing career had largely taken him away from formal education, causing him to never graduate from high school (though credible testimony showed he was highly intelligent). As he approached legal adulthood, he craved more independence in his career and in his personal life.

He began to plan a solo career and started to assemble the advisers who would help him manage his wealth and later his estate. He wanted unconflicted advice in the negotiations for any album and related tour, and the business opportunities that he correctly saw would open for him. He met and then retained a twenty-nine-year-old music lawyer named John Branca, and Jackson's already successful career began its ascent to unprecedented heights: By the time he died, Jackson's nine albums set world-record sales.

A. Off the

Wall

The first of those albums was Off the Wall, which Jackson released in 1979. It had ten previously unreleased songs, of which Jackson wrote two, “Don't Stop 'Til You Get Enough” and “Working Day and Night,” and cowrote a third, “Get [*9] on the Floor.” It was quite successful, and two of the songs reached number 1 on the Billboard Hot 100 Chart, with another two breaking into the top 10.

Around this time Jackson had formed Mijac Music catalog — a catalog that would end up comprising the publishing rights for compositions composed by a variety of composers including, most notably, compositions written by Jackson himself. Beginning in June 1980 Warner Bros. Music² became the administrator of Mijac and would continue in that role through Jackson's death.

B. Thriller

But Off the Wall was as the widow's mite to the temple treasury that was his next album — Thriller. Recorded at the end of 1982, Thriller had nine songs, of which Jackson wrote four: “Wanna Be Startin' Somethin',” “The Girl Is Mine,” “Beat It,” and “Billie Jean.” (The title song, however, was composed by Rod Temperton and not Jackson.)

Thriller was a global sensation. Seven of its nine songs became top 10 singles. The album was the number-one record in the United States for an unprecedented 37 weeks and remained in the top 10 for 80 weeks. The Recording Industry Association of America (RIAA), which certifies gold and platinum records, kept having to come up with new awards to symbolize its success, and [*10] Thriller ended up certified as 33x Multi-Platinum.³ Decades later when we tried this case, Thriller was still the top-selling album in history.

Jackson's popularity was not limited by what he did in the studio. In March 1983, during the taping of the Motown 25 television special, he took the stage to perform “Billie Jean” and debuted his iconic “moonwalk”. Motown 25 aired on May 16, 1983 to an estimated audience of 34 million viewers in those preinternet, prestreaming days. These years were also at the dawn of the music-video business, and “Michael Jackson's Thriller” debuted in December 1983 — eventually becoming, at over nine million copies, the best-selling music video ever.

By 1984 it was quite probable, as Branca would later testify, that Jackson was the most famous person in the world. He was also at the peak of his personal [*11] popularity, which was nearly universal if not quite unanimous.⁴ Branca and Jackson's other advisers had negotiated an excellent deal, and Thriller's success presented Jackson with a flood of income and unprecedented bargaining power over his recording company. Jackson's team used their power wisely, and Branca renegotiated Jackson's contract with CBS Records to transfer ownership in Jackson's master recordings⁵ to a corporation — MJJ Productions, Inc. — that Jackson himself owned.

C. Victory

Jackson remained on this extraordinarily high plateau for a couple more years. Later in 1984 The Jacksons released their final album, Victory. The album contained eight previously unreleased songs, three of which Jackson cowrote. The group went on the Victory Tour, promoted by Chuck Sullivan — the then-owner of [*12] the Boston (now the New England) Patriots football team and its stadium — and sponsored by Pepsi-Cola Company. This tour itself grossed more than $70 million, and each brother netted $6 million. Jackson donated his entire share to charity.

Jackson's personal fame meant that he received numerous requests for merchandising licenses for his “image and likeness.” To handle these requests, Triumph International, Inc. — an S corporation with Jackson as its sole shareholder — was incorporated in March 1984. In July 1984 Triumph entered into a five-year licensing agreement with Sullivan's company, Entertainment Properties, for the use of Jackson's image and likeness on a variety of products, including a line of clothing and fragrances. Under the agreement, Jackson was to receive $18 million upfront, with a potential total of $28 million. Jackson, however, was the only winner in this deal. He got the $18 million, but the merchandise didn't sell, and the deal ultimately proved disastrous for the Sullivan family.

Jackson recognized the other opportunities that his financial success presented. He sought the advice of Branca and his team, and he made two decisions that are important to this case.

The first was to buy up copyrights in musical compositions of other artists and music catalogs. After making a number of smaller purchases, Branca told [*13] Jackson that the ATV Music Publishing Catalog was for sale. The special asset in this catalog was at least 175 Beatles songs written by Paul McCartney and John Lennon. Jackson ran the idea by his investment committee, many of whom felt the asking price was too high. Jackson was also concerned about buying what he viewed as the creative property of a friend. So, before bidding on ATV, he told Branca to call Yoko Ono and Paul McCartney, and they both told him that they weren't interested. Jackson decided to go ahead with the deal and in May 1985 he bought the catalog for $47.5 million.

The money also let Jackson begin to indulge some of his eccentricities — he purchased a chimpanzee that he took along on part of a tour, installed a hyperbaric sleeping chamber, and placed a bid to buy the bones of a 19th-century medical curiosity. By the mid-'80s tabloid references to him as “Wacko Jacko” started to emerge.

D. Bad

At the time, these eccentricities seemed to have a minimal effect on Jackson's career as an artist and performer. He continued to release new material. His next solo album, Bad, came out in 1987. It had ten previously unreleased songs, of which he wrote eight. A 2001 reissue of Bad had another two songs he himself wrote, and a third — “Todo Mi Amor Eres Tú” (a Spanish language version [*14] of “I Just Can't Stop Loving You”) — for which he was a 75% composer. Bad itself became the first album ever to produce five consecutive singles to reach number 1 on the Billboard Hot 100 Chart.

Bad also pushed Jackson into his first solo tour in September 1987. The Bad World Tour was an international success; it grossed $125 million on 123 shows and drew 4.4 million fans — the most of any tour to that time.

Jackson had begun to extend his brand in other profitable ways. As it had for the Victory Tour, Pepsi again became the tour sponsor, this time paying Jackson $10-$15 million. But Pepsi wanted more than Jackson's picture on its cans, and Jackson agreed to film two Pepsi commercials and create an advertising jingle for Pepsi that featured the song “Bad”. He then packaged his recordings and performances into book sales. His autobiography Moonwalk reached the top of the New York Times bestseller list in April 1988. He starred in a short Disney film as “Captain EO”. Jackson again put his riches into valuable assets. In the latter part of 1988, Branca negotiated the purchase of the Sycamore Valley Ranch — later known as the Neverland Ranch — in Santa Barbara County, California for Jackson.

At the decade's end Jackson received the Heritage Award for Career Achievement at the Soul Train Awards. His friend Elizabeth Taylor anointed — and [*15] perhaps immortalized — him as the “true King of Pop, Rock, and Soul.” Jackson shortened the epithet to “The King of Pop.”

His fame and fortune was not quite at Thriller levels, but the money kept flowing. After the Bad World Tour ended, Jackson signed with L.A. Gear to endorse a line of cobranded sneakers for $4.5 million in cash and $3 million in L.A. Gear stock. He expected to release the sneaker line to coincide with his next album, Dangerous. But when that album was delayed, L.A. Gear nevertheless marketed the sneakers. Sales proved so bad that L.A. Gear's stock price tanked. The company responded by suing Jackson for $10 million in damages for fraud and breach of contract, and the case settled for an undisclosed amount.

E. Dangerous

In hindsight, Jackson's eccentricities and these relatively small business reversals were omens. But at the very start of the '90s, all still seemed well. In 1991, his wholly owned company, MJJ Ventures, Inc., signed a joint venture agreement with Sony that enabled him to distribute his recordings and videos through Sony Software, Inc. That same year he released his next album, Dangerous. It contained 14 songs, of which he wrote 3, “Heal the World,” “Who [*16] Is It,” and “Will You Be There,”⁶ and cowrote nine others. The album debuted at number 1 on the Billboard album chart and remained in the top 10 for more than a year, with sales of more than 30 million worldwide.

To promote the album, Jackson embarked on the Dangerous World Tour in June of 1992. Pepsi again signed on as sponsor — this time for a reported $15 million. Through Triumph, Jackson signed with Winterland Productions for a tour-merchandising agreement. The agreement gave Winterland the exclusive right to use “the name, symbols, emblems, designs, trademarks, service marks and/or copyrights in graphic designs, likenesses and visual representations of * * * MICHAEL JACKSON” in connection with the manufacture and sale of posters, paper products, and “upper body garments * * * in, around and at each concert site” where Jackson appeared as part of the Dangerous World Tour. This netted Jackson $5.2 million in advances.

Jackson released a concert program about this tour on HBO, and later on DVD, called Live From Bucharest. This generated some revenue — “enough to cover the production costs” — but failed to “generate a significant amount of income.” Jackson quickly rebounded with a halftime performance at Super Bowl [*17] XXVII in January 1993 that received the highest Super Bowl ratings since 1987.

The Dangerous World Tour, however, was to be his last in the United States.

III. The Fall

In the summer of 1993, while Jackson was still on tour, the family of 13-year-old Jordan Chandler sued Jackson for torts that included sexual battery and seduction. The allegations exploded into the press. Jackson denied any wrongdoing and he was not charged with child molestation or any other crime as a result, although a criminal investigation did begin. The allegations had lasting effects. By November Jackson canceled the remainder of his Dangerous World Tour and Pepsi cut all ties with him. A confidential settlement was eventually reached the following year. It was widely reported that the criminal case then died for want of a cooperating witness. See generally 1997 Cal. Legis. Serv. Ch. 18 (S.B. 115) (West); Gerald F. Uelmen, “Jackson Faces Tougher Laws”, L.A. Times (Nov. 23, 2003); Jim Newton, “Jackson Not Charged but Not Absolved”, L.A. Times (Sept. 22, 1994).

Jackson continued to create music and tour, albeit only internationally. In 1995 Sony released his next album, a double-disk set, HIStory: Past, Present and [*18] Future Book I. Disk 1 contained previously released songs, while Disk 2 contained 15 original songs, 7 of which Jackson wrote and 5 others which he cowrote. The album sold 20 million copies — nowhere near Thriller numbers, but still the best selling double-disk set of all time.

Jackson launched a 13-month tour to support the album in September 1996, but none of the dates was in the United States. One could begin to see a growing fissure between Jackson's popularity as a singer and his popularity as a merchandising brand. Despite the excellent album and ticket sales — the tour grossed $165 million — it went unsponsored. After Chandler's allegations against Jackson, it appeared that companies weren't interested in associating their brands with his.

The tour did have one merchandising agreement. Triumph granted Sony Signatures “the sole and exclusive right and license to utilize the Licensed Marks in connection with the manufacture and [distribution]” of merchandise, with “Licensed Marks” defined as “name(s), symbols, logos, trademarks, designs, likenesses and/or images of [Jackson].” Sony Signatures paid a $6 million advance, and Jackson also got nominal amounts from a few other licensees. Sales of tour merchandise were, however, significantly less than the advance, and Jackson had to repay Sony Signatures nearly $4 million.

[*19] That year — 1995 — also marked the beginning of the financial pressures that would ultimately come close to crushing Jackson towards the end of his life. Several of his advisers recommended that he sell his ATV catalog to Sony. Jackson agreed in part and sent Branca to negotiate a merger of the ATV music catalog with Sony's music-publishing business. In November 1995 he signed an agreement with Sony Music Publishing Company and its affiliates to form Sony/ATV Music Publishing Company, LLC (Sony/ATV).

Jackson and Sony each received half of Sony/ATV. Sony paid Jackson $115 million as an equalizing payment.⁷ Sony also promised to pay him $32.5 million over the next five years. The agreement provided that Sony and Jackson could each appoint an equal number of board members to represent their interests and that the board had the power to approve various “major decisions.” Apart from these major decisions, however, “[t]he overall business, operations and tax, accounting, financial and other affairs” of Sony/ATV was to be “exclusively managed” by Sony.

The primary purposes of the 1995 agreement were “(a) to own and exploit [Sony/ATV]'s [a]ssets and to collect income derived therefrom, [and] (b) to [*20] acquire and/or administer additional music publishing catalogs and to collect income derived therefrom.” It was also the stated intent of Sony and Jackson to “actively expand [Sony/ATV] by future acquisitions.” The deal was by all accounts a good one for both parties and left Jackson with both nearly equal power over the company and a considerably enhanced pile of cash. But he gave up complete control of one of his most valuable assets.

In 1997 Jackson released Blood on the Dance Floor: HIStory in the Mix. The album had eight remixed tracks and only five new songs, of which Jackson wrote one and cowrote the other four. Sales were not in the same league as those of his original albums but it still sold more than 11 million copies worldwide, which made it the most successful remix album of all time. There would, however, be no associated tour. But a problem began to arise — even with cash flowing in from the Sony/ATV deal and his music, Jackson's spending was starting to outpace his income, and he began to borrow significant sums against his share of Sony/ATV. By the end of 1998, his interest was burdened by $140 million in debt to Bank of America.

IV. The Collapse

In the next two years his borrowing against the Sony/ATV interest had increased to $185 million. It seemed that Jackson was about to become a real-life [*21] Mike Campbell — bankrupt in two ways, first gradually, then suddenly. See Ernest Hemingway, The Sun Also Rises 72 (1926). It did not help his finances that he became increasingly reclusive. His public performances were limited to two tribute concerts at Madison Square Garden to mark the thirtieth anniversary of the recording of his first studio solo album. They were watched by an enormous television audience, but for most of the show Jackson sat in what one reviewer called a “royal viewing box,” and any favorable publicity was soon overwhelmed by the September 11 attacks days later.

These concerts were to be his last.

Although his tour income vanished, Jackson did continue to release records. Invincible came out in 2001, with 16 new songs of which Jackson wrote 2 and cowrote 12. But this album had an extremely large production budget of $30 million and a promotional campaign that cost $25 million. With costs like this and with Jackson either unwilling or unable to tour in support of it, Sony didn't want to distribute the album and refused to renew his recording contract. Epic Records picked up distribution, but sales amounted to only 4.5 million albums in the first year. This would have been an astounding number for an ordinary star; for Jackson they were a disappointment. And he wouldn't release another wholly original album as long as he lived.

[*22] A. The Bashir Documentary and Criminal Case

Around the time Invincible shipped to record stores, Jackson agreed to allow British filmmaker Martin Bashir to spend the better part of a year with him. This decision proved disastrous. Bashir and his crew — with Jackson's permission — followed him around Neverland to make a documentary called Living With Michael Jackson. The documentary, released in 2002, focused on the most unusual parts of Jackson's life — his skin bleaching, his plastic surgery, and his relationships with young boys.

With the release of this documentary, an even larger fraction of the public began to view Jackson as a pederast. This included local law enforcement. In the summer of 2003 the Santa Barbara County District Attorney's Office began a criminal investigation into allegations that Jackson had molested another 13-year-old, this one named Gavin Arvizo. Jackson surrendered to the Santa Barbara County Sheriff's Department in November 2003 and faced seven counts of child sexual abuse and two counts of administering an intoxicating agent. He pleaded not guilty and was tried in January 2005. In June he was acquitted of all counts.

[*23] B. Financial Peril

Acquittal did not rehabilitate his reputation. And this ordeal did nothing to stanch the outflow of his wealth. He continued to spend great gobs of money in excess of his now shrinking income. To keep pace, Jackson took out additional loans secured by his assets. The terms of these loans became ever more onerous because his acquittal was not an exoneration and fewer banks were willing to even consider doing business with him after the criminal trial. One very important instance of this happened in 2005 when Bank of America cut ties for fear of its own reputation. It did this with a sale of Jackson's debt to Fortress Capital Corporation. Fortress was a distressed-debt hedge fund that lent money on terms that aimed less at repayment and more at ultimately gaining control of the underlying collateral — which now included Neverland.

Fortress didn't care that much about Jackson's reputation. It seemed happy, in a usurious way, to increase Jackson's debt to around $270 million.⁸ This had two effects. The first was that Sony required him to agree to amendments to the Sony/ATV operating agreement before it would allow him to further encumber his interest. These amendments included a grant to Sony of an option to buy 50% of [*24] Jackson's interest at a value capped at a maximum formula price determined as of March 2006. The second was that Jackson's annual interest payments to service the loan increased to around $15 million — well in excess of his annual distribution from Sony/ATV.

It looked like Jackson was moonwalking from the “gradually” to the “suddenly” part of going bankrupt as his financial position kept advancing backward. The difference between the interest he owed and the income he made forced him into another refinancing in 2007. Jackson got this done in December with a pledge of additional collateral. Now his primary loan was secured not just by his interest in Sony/ATV but by his Mijac Music catalog.

The refinancing was not easy. Jackson's team contacted dozens of banks, and in the end Barclays Bank and Deutsche Bank offered to fund a loan secured by Sony/ATV, and HSBC and Plainfield Asset Management offered two separate loans secured by Mijac Music. Jackson eventually made the following deals:

• Barclays refinanced $300 million of the debt, in the form of bonds fully guaranteed by Sony and issued by a new entity, New Horizon Trust II, a bankruptcy trust,⁹ which held Jackson's economic interest in Sony/ATV. All of Jackson's annual distributions from Sony/ATV were redirected to an interest reserve for this New Horizon Trust II debt. None of the loan proceeds went to him personally. [*25]

• HSBC lent an additional $30 million secured by Mijac Music through New Horizon Trust III, another bankruptcy trust. In addition to Mijac Music, Jackson assigned to NHT III his writer's share of performance royalties from Broadcast Music, Inc. (BMI) — after BMI's recoupment of any advances paid to Jackson — and redirected them through the trust to HSBC to pay the interest on this loan.

• Plainfield Asset Management lent $40 million, secured by a second lien on Mijac Music through New Horizon Trust III. This loan did not require current payments of either principal or interest, but interest accrued at 16.5% annually.¹⁰ This interest was added to principal that would be due on the maturity date, December 31, 2011.

These deals were, under the circumstances, an achievement. But Branca was on his way out. Since 2002 his relationship with Jackson had deteriorated, though Branca had continued to represent Jackson through the 2005 criminal trial. Jackson's eccentricity by now extended to his choice of advisers, and Branca felt that this new team was filled with incompetent people who did not have Jackson's “best interests at heart.” Branca formally resigned as Jackson's attorney by 2006.

C. Jackson Leaves Neverland

Jackson also had left the country. After his acquittal, he and his children moved to Bahrain to stay as guests of Sheikh Abdullah, son of the Bahraini king. From 2006 to 2008 Jackson and his family shuffled among various countries. He [*26] returned to the United States in 2008 and lived in a rented house in Las Vegas. By the fall of 2008 he moved once again, this time to a rented house in Los Angeles.

Jackson had come home to yet more financial trouble. By early 2008, Fortress was edging towards foreclosure on Neverland. Since Jackson did not want to lose Neverland, his brother Jermaine approached a man named Dr. Tohme Tohme,¹¹ who represented Colony Capital, a private equity firm headed by Tom Barrack. Tohme received from Colony Capital $20,000 per month and a part of any deals he arranged for that company. He agreed to help Jackson avoid foreclosure.

Colony Capital did eventually agree to buy the Neverland mortgage from Fortress. This avoided foreclosure, but Jackson was forced to contribute the Neverland Ranch to Sycamore Valley Ranch Company, LLC, and remove all of his personal property.

This was a difficult deal for Jackson to accept, but there was at least one sign of relief. In February 2008 Thriller 25 was released to commemorate the twenty-fifth anniversary of the original album's release. Thriller 25 was [*27] particularly successful and led to a substantial, but temporary, increase in Mijac royalties. That increase, however, didn't materialize until the first half of 2009 as a result of the usual delay in the reporting and payment of music royalties.

V. The Unconsummated Comeback

After this crisis faded, Jackson hired Tohme as his manager. It was an expensive arrangement, as he signed a contract in July 2008 in which he promised to pay Tohme a fixed monthly fee of $35,000 plus 15% of all gross compensation that he received as a result of Tohme's services.¹²

Tohme immediately set out to find endorsement, licensing, and personal-services opportunities for Jackson. There was a discernible lack of professionalism in some of his efforts. He spoke, for example, with Jack Wishna, the CEO of an entertainment consulting firm named CPAmerica, about creating a Jackson-themed Cirque du Soleil show. Wishna told Tohme that he worked for Cirque du Soleil, but Tohme did not do even a simple background check. It turned out that Wishna had nothing to do with Cirque du Soleil, but had called Daniel Lamarre, its CEO, and claimed to have the rights to do a show based on Jackson's [*28] music, which he did not. No deal ever came from these discussions, and Tohme never spoke to any actual Cirque representative.

Tohme also contacted Andy Heyward about creating an animated children's show around the “Thriller” song and video. Heyward is a writer and director who has spent his career producing children's entertainment, including Inspector Gadget, Dennis the Menace, and Ghostbusters. The contemplated “Thriller” series would have incorporated “several signatures from the [Thriller music] video,” including ghouls, zombies, and other creatures. Heyward, however, wanted nothing to do with making Jackson a character or even using Jackson's image and likeness.

In the end, there was no series. An email from June 2009 from Heyward to Tohme states the obvious reason: “[A]s sweet as [Jackson] is, the parade of 'unusual' personalities in his life, leaves me very nervous. I have spent 25 years in kids entertainment, building a reputation, and there is no amount of money or riches that I would jeopardize that for.” Heyward reiterated this sentiment in his altogether credible testimony, where he described Jackson's image and likeness in children's entertainment as “candidly toxic.” He and his company “didn't want to have any signature which would associate [Jackson] with this brand.”

[*29] Tohme wasn't the only person looking to find potential marketing opportunities for Jackson. Jackson also charged Peter Lopez, one of Branca's successors, with finding endorsement deals. Lopez reached out to Nederlander Presentations, the famous Broadway production company, about doing a Jackson-themed show. These discussions even led to a deal memorandum in October 2008 that would have given Nederlander the exclusive right to create a musical based on compositions performed by Jackson in the Off the Wall and Thriller albums. The deal was never consummated while he was alive.

A. This Is It Tour

By 2008 Jackson was sufficiently strapped for cash that he was willing to tour again. Barrack (the businessman who had helped Jackson avoid losing Neverland) called Phil Anschutz, the owner of Anschutz Entertainment Group (AEG), to tell him that Jackson wanted to work again and asked whether Randy Phillips — the CEO of AEG Live, a concert promoter and subsidiary of AEG — would be interested in organizing a concert tour. Negotiations began soon after.

Phillips ultimately decided it was best to start a tour in London. AEG had recently built the O2 Arena there, and Phillips also felt that the best way to rehabilitate Jackson's image was to start the tour abroad. He believed that people [*30] overseas were less judgmental and more forgiving — or, perhaps, less sunk in the flood of sensational news reports from American media. If all went well, he thought, the tour might end with dates back in the United States.

These negotiations were prolonged, but Jackson and his advisers stuck with them because his cashflow problems were severe and obvious. The deal was made in January 2009 and comprised two agreements. In one Jackson promised to perform services for a concert series to be called “This Is It” and in the other to develop up to three feature films. The concert agreement gave AEG Live the exclusive right to:

• promote the shows;

• manufacture and sell, or arrange for others to manufacture and sell, Jackson merchandise at the shows; and

• solicit sponsors.

It also granted AEG Live the nonexclusive right to use Jackson's image and likeness in connection with the exercise of its contract rights. Jackson was in turn to receive several advances including an initial one of $5 million, and a $100,000-per-month advance to pay the rent on a home in the Holmby Hills area of Los Angeles for 12 months. Each of these advances was to be recouped from tour proceeds.

[*31] Jackson secured AEG Live's promise to film rehearsals to help him prepare, plan, and develop his concert performances. The company bought two handheld cameras and hired two cameramen to record Jackson's rehearsals. This was not unusual — Jackson often had videos shot of his rehearsals to perfect his public performances. But these videos were not intended for public consumption as Jackson never sang full out during his rehearsals, in an effort to not harm his voice. The record contains no indication that this footage was ever intended for release by anyone or even that anyone other than Jackson himself ever expected to look at it.

AEG Live also had the right to film Jackson in concert, and it intended to release a professionally shot two-disc concert DVD, with one disc that captured Jackson's concert performances and another that showed behind-the-scenes activity and interviews. AEG Live got this more polished production underway even while Jackson was still putting the show together — there was professionally shot footage of dancers as they auditioned for the tour. But the production's other costs started to rise sharply, and AEG Live soon realized that the $20,000 daily cost of shooting this footage was not sustainable.

[*32] 1. The This Is It Tour Is Announced

Jackson publicly announced the This Is It concert tour at a press conference in London in March 2009, and the tour was set to begin in July. When tickets for 10 concerts all sold out rapidly, AEG Live scheduled an additional 40 concerts. These too sold out. Yet even in London, one can see the pattern that had long since marked Jackson back home — a great appreciation for his music and performance, but little for his personal reputation. Despite the near instant sellout of dozens of performances, AEG Live was utterly unable to find a tour sponsor. Potential sponsors did not want to tie their own reputations to Jackson's, and there was also a fear that Jackson's troubling behavior would again flare up and cause the tour to be canceled or cut short.

2. Discussions for a Tour Merchandising Agreement

Though unable to get a sponsor for the tour, AEG Live still thought it likely that it could find a deal for tour-related merchandise. Those familiar with the industry believe that people who buy tour “merch” do so as a souvenir of their attending a concert; those who buy branded merchandise in stores do so out of admiration for identification with the person whose brand they are buying. AEG began discussions with Bravado International Merchandising Services, Inc. to become the exclusive merchandiser for the This Is It concert tour. Music [*33] merchandising consists of signing the rights for an artist's image and likeness, which the merchandiser will then use on different types of merchandise. As of 2009 Bravado was one of the leading worldwide merchandisers and had worked with a large number of well-known acts including the Rolling Stones, Metallica, Guns N' Roses, Rihanna, and Justin Bieber.

Bravado met with Jackson twice and showed him about 300 designs for tour merchandise. Jackson approved 295. Even with all this preliminary work and Jackson's cooperative attitude, Bravado refused to enter into a tour-merchandising agreement until the tour began for fear that Jackson wouldn't perform.

No deal was ever reached.

B. Jackson Cleans House

As the tour's debut neared, Jackson became more focused on his craft and more sensible in his choice of advisers. In the spring of 2009, he ousted Tohme and rehired Frank DiLeo, who had been his personal manager during the height of his career in the 1980s. Jackson also hired Michael Kane as his business manager. This reboot became complete when DiLeo set up a meeting with Branca who met with Jackson on June 17. They discussed ways to commercialize Jackson's image and likeness.

[*34] It seemed as if the chaos that was Jackson's financial life was perhaps beginning to drain away. There had been no further charges of abuse, and his engagement in choreographing and arranging the concert series seemed to have brought a much needed discipline to his personal life.

Eight days later he was dead.

VI. Jackson's Death

On June 25, less than three weeks before the tour was to begin, Jackson was killed by an injection of propofol and benzodiazepine administered by his personal physician, Dr. Conrad Murray. His death was ruled a homicide.

At the time of his death, each of the three assets that we have to value was distressed. His image and likeness was not producing any noticeable income, and he had not even been able to contract for tour merchandise. His interest in Sony/ATV secured $303 million in loans, with maturity dates less than 18 months away. Although Jackson was guaranteed $11 million in annual distributions under the Sony/ATV agreement through September 2011 — after which distributions were at Sony's discretion — the average annual interest payments on the loans were well over $17 million. Jackson's interest in Mijac secured over $72 million in debt. And he had incurred substantial new debts — remember that the advances from AEG Live were recoupable. Without the possibility of a tour, those debts [*35] somehow had to be paid. AEG Live claimed the debt amounted to approximately $36 million, and it wanted its money immediately. Various other creditors began circling around Jackson's assets. Business manager Kane added it all up and identified $450 million in “easily identifiable debt” secured by Sony/ATV, Mijac, the Neverland Ranch, and the Jackson family's Hayvenhurst home in Encino, California.

A. The Executors Take Charge

Jackson had put together a new and much more competent team of advisers just in time. They did not know precisely how much debt there was or what assets might be available. But in a remarkable and somewhat coldblooded way, the team put out an APB. Soon the hospital where Jackson's body lay surrounded by much of his family had another room filled with a haphazard team of intellectual-property lawyers and advisers — so many they might have outnumbered Jackson's family. Once there, they immediately began to discuss the administration of the Estate and how to protect his image and likeness. But it was unclear who would administer the Estate — Jackson was only 50, his death was unexpected, and nobody even knew if there was a will. Then Tohme added to the confusion when he claimed he held something he called a postdeath power-of-attorney and tried to reenter the inner circle.

[*36] A search was made, and a will was found.¹³ This will was from 2002, before the dispersion of Jackson's original team of advisers, and it named Branca as a coexecutor with John McClain, Jackson's childhood friend and a music producer. McClain, though still coexecutor at the time of trial, was himself suffering from some reverses to his health, and Branca took on most of the work. His first goal was to avoid foreclosure on Jackson's assets. He met with Phillips upstairs at Mr. Chow Restaurant to discuss how to “make as much money from whatever [assets the Estate] had.” This meeting did not feature discussions about making a Cirque du Soleil show or a movie based on Jackson's rehearsal footage, which Branca didn't know existed. It was not until a week after the dinner at Mr. Chow that — during a lunch between Jim Gianopulos (chairman of Fox Film Studios) and Phillips — the idea of making a film based on the This Is It rehearsal footage first popped up. Gianopulos said he was interested, but no one had yet seen the rehearsal footage.

[*37] B. The Memorial Service

What Jackson had created during his lifetime was now fixed, and it was to the considerable benefit of the Estate that he was no longer able to get in the way of the rational profit maximizers who were now in control. And nearly everyone involved in these early days after Jackson's death turned out to be accomplished in the business side of the entertainment business. As crass as it might have seemed to Jackson's more sentimental fans, the business began almost immediately. A popular star of Jackson's stature from a family as large as his would understandably receive a memorial service. Jackson's was produced by AEG and was held in the Staples Center in Los Angeles on July 7, almost two weeks after his death. It was an extraordinary production that featured numerous celebrity appearances and singing performances. Jackson was eulogized by both his daughter Paris and Berry Gordy. Paris endearingly described Jackson as one of the world's greatest dads, and Gordy called Jackson the “greatest entertainer that ever lived.” Gordy's statement became the centerpiece of the Estate's rebranding of Jackson. Video of the service was quickly copyrighted, and it produces income for the Estate to this day.

[*38] C. The Motion Picture: Michael Jackson's This Is It

The same week as the memorial service, AEG filed to register This Is It tour rehearsal footage with the U.S. Copyright Office. AEG did not have any concrete plans on what to do with the footage because no one had ever seen it and so no one knew its quality or content. But after Phillips met with Gianopulos, AEG employees began to review the footage with an eye toward making it into a movie. They did not start with much hope: The raw material looked as if it might be too crude to ever be useful. As Phillips credibly testified: “We kind of realized that we might have enough to make a movie, with some kind of narrative.”

Notwithstanding this uncertainty, AEG put together a 90-second “sizzle reel”¹⁴ of footage to show to movie studios. The sizzle reel was shown to several of them, including Paramount, Columbia Pictures, Fox, Warner, and Universal. AEG's corporate instincts were correct — the sizzle reel ignited a bidding war. Columbia (a subsidiary of Sony) won the war, “not because it was necessarily the best bid, but [because] they also controlled the music rights, the underlying music rights and stuff like that.”

[*39] But who owned the footage? AEG had paid for it, but it was for Jackson's own use; and it was his image and his embryonic performance that it captured and that made it valuable. Branca suggested to AEG, with some vigor, that he thought that meant the footage belonged to the Estate. Both sides realized that time was of the essence, and both stood to lose if they started a fight over who owned what and when. Contentious negotiations ensued with each side expressing significant doubt over whether a deal could be reached. On July 27, 2009, however, AEG, Columbia, and the Estate negotiated a term sheet that set out their basic agreement about the use of the rehearsal footage to develop what became the film Michael Jackson's This Is It.

Branca petitioned the probate court to approve the agreement, lest it be void ab initio. In early August, progress again ground to a halt when Jackson's mother filed a response to Branca's petition in which she reiterated the argument that AEG did not own the footage, and she added that her son never would have wanted such unfiltered footage released. The court set a hearing for later that month, and everyone knew that if the Estate lost, the film wouldn't be produced.

The probate court granted the petition in its entirety. With this approval, production could proceed, but it remained unclear whether the quantity and quality of the footage would be sufficient to make a full-length film. As Phillips testified, [*40] the footage was “[b]arely, just barely” enough. Poor lighting and audio were just part of the problem. The Estate also needed licenses from the music publishers who held copyrights to Jackson's songs. This was particularly true of the song “Thriller”, the rights to which were held by its composer, Temperton.

This team did its work well. On August 21, 2009, Sony Pictures and Sony Music Entertainment announced the release of Michael Jackson's This Is It. The Estate promoted the film with a poster of Jackson's silhouette filled with a collage of scenes from the movie. The only trademark of the Estate used on the poster was Jackson's dancing feet logo in the lower left-hand corner.

The movie was released in October 2009. It was accompanied by a soundtrack released the same day. The album, also titled This Is It, contained 16 songs. The only previously unreleased song on the album — titled “This Is It” — was based on a demo of a song recorded in the 1980s that Jackson had cowritten with Paul Anka. The demo was badly distorted and required significant effort and expense to put into releasable form.

It all paid off. The success of the film was unprecedented. As of July 2011 the movie had generated cumulative gross receipts of over $240 million, and it became the highest worldwide grossing concert documentary ever made. It was also, at the time of its release, the only concert documentary to consist entirely of [*41] rehearsal footage. The Estate had negotiated from a relatively weak position and had to rely on Columbia to put in money up front to get the movie done. The distribution deal allowed Columbia to recoup this advance first. But the movie was so profitable that by April 2010 the Estate began to receive a little money. Then it began to receive a lot of money — $45 million by the end of July.¹⁵

Michael Jackson's This Is It showed fans a side of Jackson previously unseen and it turned out to be an important part of the Estate's rebranding — shifting the public's attention to his music and away from his personal life. It “focuse[d] on [Jackson], the artist, and his genius and his talent as an artist and * * * as a human being in the way he dealt with the members of the band and the dancers.”

D. Cirque duSoleil

Despite Tohme's testimony, which we do not credit, Cirque du Soleil had not considered doing a Jackson-themed show while Jackson was alive. The Estate itself hadn't considered the possibility of such a show until Rene Angelil — Celine Dion's late husband and manager — called Branca a couple months after Jackson's [*42] death and said “that Cirque would be interested in talking about the possibility of creating some kind of show based on [Jackson's] music.”

“[E]ncouraged and emboldened” by the success of Michael Jackson's This Is It, Branca began talks with Cirque du Soleil to develop a Jackson-themed show in September 2009. Cirque's chairman, Guy Laliberté, told Branca that he wanted to create a worldwide traveling show, while Branca told Laliberté he wanted a resident show in Las Vegas where Cirque had created other shows like The Beatles Love show. Both got what they wanted, and Cirque ended up planning two Jackson-inspired shows, IMMORTAL and ONE.

Despite Jackson's tarnished personal image, “Cirque believed that [Jackson's] music had not been affected by the litigations in which [Jackson] had been involved” and that “[Jackson's] music was still very popular in people's minds.” And so, in February 2010, the Estate returned to probate court with a petition to approve “an agreement * * * with Créations Méandres Inc. for the presentation of a live show based on [Jackson's] music and songs.”

In the ensuing agreement for the Las Vegas resident show, the Estate got “the right to approve the use of all Michael Jackson's images, likenesses, manner in which Michael Jackson is depicted, and other portrayal[s] to be included in the [s]how.” It also set forth the basic financial terms. Cirque agreed to fund up to [*43] $46 million in development and operating costs for which it would be reimbursed in amounts that were the same as or less than it had received for previous shows based on the Beatles and Elvis Presley. The Estate and Cirque each owned 50% of the rights in the show, with 80% of profits being used to repay Cirque its investment (with interest), and the remaining 20% to be split equally between them. After Cirque was reimbursed for its costs, the profits were to be shared equally between the Estate and Cirque.

In October 2011 the touring show, Michael Jackson: THE IMMORTAL World Tour premiered. It cost about $50 million to produce and was funded exclusively by Cirque. Cirque at first agreed to license Jackson's image and likeness in THE IMMORTAL at 3.5% of the gross. But, as described by Branca, the license was a “land grab” by the Estate, which Cirque ultimately reduced and then terminated “because the cost of the show [w]as so prohibitive that [Cirque] couldn't continue to pay it.” Branca thought THE IMMORTAL earned the Estate $20 to $25 million, which included the income streams from image and likeness, profits, masters, and publishing.

In June 2013, the Las Vegas show, Michael Jackson: ONE, premiered. It combined acrobatics, dance, and visuals — immersing the audience in the world of Jackson's music. The cast included four Cirque performers each of whom was [*44] given one of Jackson's iconic items — his white glove, penny loafers, fedora hat, and shades — and they appeared as “misfits” who set out on a journey into Jackson's music. Despite the license to use, among other rights, Jackson's image and likeness, we find that Cirque refused to pay anything specifically for their use.

E. General Merchandising Agreement With Bravado

Bravado's CEO, Tom Bennett, credibly testified that Bravado would not have done a nontour, general merchandising deal for Jackson's image and likeness before he died for “any meaningful money” because there was simply “no demand.” But Bennett felt that Jackson's death presented new opportunities. In his experience “when any big celebrity dies, there's an immediate desire in the marketplace for memorabilia merchandise, and you never know how long that's going to last.” To project the profitability of a merchandising deal, Bravado contacted retailers to gauge their interest. Walmart had none, and it believed “that the Michael Jackson brand was not something that * * * was consistent with their consumer base, because of the alleged allegations of child molestation.” Bravado received a “lukewarm response” from smaller retailers.

Bravado then decided to move forward with a merchandising agreement with AEG Live, which held the rights to the 295 images and products that Jackson approved before he died. They reached an agreement in early July 2009 that gave [*45] Bravado the exclusive right to use Jackson's “names or sobriquets, symbols, emblems, logos, designs, likenesses, visual representations, service marks, and/or trademarks * * * in connection with the manufacture, advertisement, merchandising, promotion, distribution, and sale of mutually approved merchandise.” Bravado paid AEG Live a $5 million recoupable advance for this right, and then immediately began manufacturing the merchandise.

Sometime after that, Bravado entered into an agreement with the Estate itself. Compared to its agreement with AEG, the agreement with the Estate granted Bravado “a broader body of rights” including “licensing out [Jackson's] name and image.” But the Estate reserved a number of rights to specific categories of Jackson's image and likeness, including, but not limited to, music-based products, licensing of music compositions, video games, and audiovisual works. For these rights, Bravado paid the Estate $10 million in a recoupable advance.

As it turned out, selling Jackson-themed merchandise “was not the bonanza that [Bravado] thought it may have been” as licensees proved much harder to sell than expected. It was only after five years, and a nontraditional merchandising arrangement with a slot-machine company, that Bravado recouped its advances. This was far longer than the company had expected.

[*46] F. Miscellaneous Agreements

The Estate built upon its success by entering into a number of other, smaller deals. One was an agreement at the end of 2009 to create an online platform for Michael Jackson's Official Virtual World Massively Multiplayer Online Game. The agreement provided for a 15% royalty and recoupable advances of $4 million within the first year, $3.5 million in the second year, and $2.5 million in the third year.

In March 2010 the Estate made a second video-game licensing deal. Under this agreement, the Estate was to be paid four $1-million installments, which were fully recoupable against royalties.

G. Posthumous Albums and the Hunt for Unreleased Songs

After Jackson's death, Sony engaged in an extensive search for unreleased songs to evaluate for possible future release. Jackson was not working on any album when he died and had not released any album containing new material since 2001. He was, however, known to over-record songs on his albums. He kept these unreleased recordings in his personal vaults.

Sony's corporate spelunkers crawled through these vaults and found 7,000 to 10,000 pieces of tape. These were mostly tailings and very little pay dirt. There were only 2 completed and unreleased recordings and approximately 25-30 full [*47] vocals with some music. The Estate has confirmed a total of 83 songs — fragments of lyrics, tunes, and vocals — that were unreleased at the time of Jackson's death.

They also found that there was often a reason for an unreleased song to remain unreleased, or a “full vocal” not to be a “song”. Sony executive John Doelp credibly testified that once a vocal was identified, Sony had “to take a step back” and ask whether it was commercially viable. Doelp described the process as follows: “[I]f it's a demo vocal, it's very possible that it's just a bad performance. There could be notes that are flat * * * [or] not well recorded. * * * [I]t could just not sound good, and then the song itself just might not be good or just not up to Michael's standards[.]”

There was some refined gold beneath the dross, so in November 2009, the Estate and Sony Music Entertainment contracted for the Estate to deliver 10 posthumous albums between October 2009 and December 2016. Of those 10 albums, however, only 2 required delivery of master recordings of previously unreleased compositions. These 2 albums required 10 to 13 songs. There was no requirement that Jackson composed these songs, only that he performed them. An additional three anniversary albums were to “contain previously unreleased [r]ecordings derived from” the songs on the original albums. (Emphasis added.) [*48] These derivations could, for example, include “outtakes, demos, [and] alternate versions.” The anniversary albums neither contemplated nor required the delivery of previously unreleased compositions.

1. Michael and Xscape

In December 2010 Sony released Michael. It had 10 previously unreleased songs, 5 of which Jackson had written or cowritten. In May 2014 Sony released Xscape. Xscape was the second posthumous album with entirely previously unreleased recordings. But while the agreement required at least 10 new songs, Xscape contained only 8, only 5 of which Jackson wrote or cowrote.

2. Bad 25

In September 2012, Sony released Bad 25, a twenty-fifth anniversary edition of the album Bad. Bad 25 was a two-disc set. Disc 1 contained the songs on the original Bad album, and disc 2 contained remixes of several old songs, as well as six previously unreleased songs.

H. Estate Sells Jackson's Interest in Sony/ATV to Sony

In 2011, Sony/ATV acquired EMI Music Publishing. With this acquisition, Sony/ATV went from the fourth largest music-publishing business to number one. Five years later, and seven years after Jackson's death, Sony and the Estate signed [*49] a deal for Sony to acquire the Estate's interest in Sony/ATV and become the 100% owner of Sony/ATV. Sony paid $750 million.

The Estate was very, very far away from the condition its executors and managers thought it was in at that first dinner in 2009 in the upper room at Mr. Chow.

VII. The Estate Prepares Its Return

After Jackson died the Estate hired the accounting firm Crowe Horwath to prepare its return. Their work began with a long list of labor-intensive chores, as Jackson's various assets needed to be inventoried, photographed, valued, and insured. The chores became more arduous when they discovered that Jackson's Tohme-led team of advisers had kept no contemporaneous books and records in the last three years of his life. Kane, Jackson's last business manager, worked as a liaison between the Estate and the Crowe Horwath team to prepare the return, and recommended appraisers to value Jackson's interests in Sony/ATV and Mijac, as well as his image and likeness.

The Estate retained Moss Adams, a large accounting and consulting firm, to value Jackson's image and likeness and his interest in Mijac. Relying entirely on the income approach to valuation, Moss Adams valued Jackson's image and likeness at $2,105 and Mijac at $70,860,000.

[*50] To value Jackson's ownership interest in Sony/ATV, the Estate selected the Salter Group, an independent financial and strategic advisory firm that specializes in valuations. The Salter Group also chose to use only the income method, which led it to value Jackson's ownership interest in Sony/ATV at $0.

Using these valuations, the Estate, on its 2009 Form 706, United States Estate (and Generation-Skipping Transfer) Tax Return, reported the value of:

• Jackson's image and likeness at $2,105;

• NHT II, which held Sony/ATV, at $0; and

• NHT III, which held Mijac, at $2,207,351.

The Estate also reported Jackson's various other assets — the most significant among them including his Hayvenhurst home; MJJ Productions, Inc.; MJJ Ventures, Inc.; and his master recordings. Jackson was, when he died, the sole shareholder of MJJ Productions and MJJ Ventures. MJJ Productions collected Jackson's mechanical royalties as a recording artist under an agreement with Sony Music, Inc., while MJJ Ventures collected Jackson's share of joint venture income under an agreement with Sony for the exploitation of Jackson's master recordings.¹⁶

[*51] VIII. The Audit

The Commissioner audited the Estate's tax return and in May 2013 issued a notice of deficiency that adjusted the Estate's reported values. We summarize it here:

Item	Adjustment
Hayvenhurst real estate	$1,425,000
MJJ Ventures, Inc.	67,393,780
Share of artist mechanical rights under Jackson 5 master recordings, and master recordings	34,299,095
Miscellaneous property	48,603,827
Image and likeness	434,261,895
New Horizon Trust II	469,005,086
New Horizon Trust III	58,478,593
Debts	12,252,591
Limitations to Schs. J & K	(713,436)
Total	1,125,006,431

This adjusted valuation led the Commissioner to conclude that the Estate had underpaid Jackson's estate tax by a shade more than $500 million. The Commissioner also determined that some of the valuations were so far off that he tacked on penalties of nearly $200 million. The Estate timely petitioned. During trial, neither the Commissioner nor the Estate entered any evidence into the record [*52] to show that the initial determination of these penalties was personally approved in writing by the immediate supervisor of the individual making the determination. The Commissioner moved after trial to reopen the record with new evidence to try to show that the initial determination of penalties was personally approved in writing by the immediate supervisor of the individual making the determination. We denied that motion.

IX. Pretrial Preparation and Stipulation

This is a very large case, and the parties reasonably asked for a lengthy pretrial phase in which they were able to settle a great many of their disputes. But they got stuck in their negotiations on the value of three of Jackson's assets:

• his image and likeness;

• NHT II, which held his 50% ownership interest in Sony/ATV; and

• NHT III, which held his ownership interest in Mijac Music.¹⁷

[*53] X. Trial

We tried this case in Los Angeles. Jackson was domiciled in California at the time of his death, and both executors were residents of the state when they filed the petition. Any appeal would presumptively go to the Ninth Circuit. See sec. 7482(b)(1)(A).

A. The Estate's Experts

The Estate retained four experts for trial:

• Mark Roesler and Jay Fishman to value Jackson's image and likeness,

• Alan Wallis to value Jackson's interest in Sony/ATV, and

• Owen Dahl to value Jackson's interest in Mijac Music.

Roesler is the founder and CEO of CMG Worldwide, Inc., an international licensing and rights-management company that specializes in representing celebrities both dead and alive, including Marilyn Monroe, James Dean, Buddy Holly, Chuck Berry, Princess Diana, and Jackie Robinson. Fishman has been a professional appraiser since 1974 and is a managing director of Financial Research Associates, a firm that provides business-valuation, forensic-accounting, and litigation-consulting services.

To do his job, Roesler projected 10 years of postdeath revenues from the exploitation of Jackson's image and likeness under California law, see Cal. Civ. [*54] Code sec. 3344.1 (West 2012), and from some associated trademarks. Fishman then used Roesler's revenue projections as the starting point for his own use of the income method to value Jackson's image and likeness. After he figured out future cashflows, Fishman discounted the stream to present value and came up with a higher value for this asset than the Estate had on its return — about $3 million instead of $2,105.

Wallis has valued businesses for over thirty years and leads the Media and Entertainment team in Ernst & Young's UK valuation practice. In that role, he values media-related intellectual property (including copyrights), and content and character licenses. Over the course of his career, he has valued approximately 100 music-publishing and master-recording catalogs.

The value of NHT II depends on the value of Jackson's interest in Sony/ATV. Wallis valued Sony/ATV using two different valuation methods — the market approach and the income approach. The first required him to compare Sony/ATV to comparable companies and comparable transactions. Wallis viewed Sony/ATV as an operating music-publishing company. This required him to calculate Sony/ATV's earnings before interest, tax, depreciation, and amortization [*55] (EBITDA)¹⁸ as a key figure in his calculations. The second method required him to project Sony/ATV's future cashflows. His basis for these projections was Sony/ATV's own internal projections as of 2009. He concluded that Sony/ATV's enterprise value was $1.1 billion. He then reduced that value by Sony/ATV's net debt to arrive at the company's equity value. Because Jackson was a 50% owner, he halved the equity value, which equaled $254 million. But remember that Jackson owed debt secured by his interest, and he had further encumbered that interest with restrictions on his power to control the joint venture or sell his stake in it. After Wallis analyzed these facts, he concluded that NHT II was worth nothing when Jackson died.

Dahl is the president and founder of the Dahl Consulting Group, a full-service valuation-consulting firm. From 1999 to 2012 Dahl was a principal at Moss Adams, where he specialized in the appraisal of intellectual property. During his career he has valued various high-profile music catalogs, including Cherry Lane Music Publishing and Holland-Dozier-Holland.

[*56] Dahl valued Jackson's interest in Mijac, and he also used the income approach. He identified five sources of income:

• Jackson's compositions which he also performed that were released before his death,

• Jackson's compositions which he didn't perform that were released before his death,

• major works by other songwriters,

• minor works by other songwriters,

• Jackson's unreleased compositions that he performed.

He then analyzed how much income each of these sources would produce. He calculated Mijac's value to be about $71 million. After he added cash on hand and subtracted the debt that Mijac secured, Dahl concluded that the fair market value of NHT III was about $2.7 million.

Each of the Estate's experts reduced the cashflows produced by the assets to reflect the tax implications to a hypothetical buyer — a process known as tax affecting. Nancy Fannon testified for the Estate about the current state of academic research on the topic and on the empirical evidence that she said proved the prospect of taxes would affect the price a prospective buyer would be willing to pay.

[*57] B. The Commissioner's Expert

1. The Valuations

The Commissioner faced this chorus of experts with a soloist. Weston Anson is the chairman of CONSOR Intellectual Asset Management, an intellectual-property consulting firm that specializes in trademark, patent, and copyright valuations. For more than 25 years, he has valued intangible assets including those of Dr. Seuss, Andy Warhol, Tupac Shakur, Audrey Hepburn, Marlon Brando, and Woody Allen.

Anson used different methods to value the three big assets that were at stake. To value Jackson's image and likeness, Anson considered five “opportunities” that he believed a hypothetical buyer could reasonably foresee at Jackson's death:

• themed attractions and products,

• branded merchandise,

• a Cirque du Soleil show,

• a film, and

• a Broadway musical.

According to Anson, image and likeness encompasses a broad bundle of allied rights, including U.S. trademarks, state or common-law trademarks, copyrights, [*58] licensing rights, endorsement rights, franchising rights, and international trademarks. Employing the income approach, Anson determined the value of Jackson's image and likeness to be $161 million.

Anson valued Jackson's interest in Sony/ATV through both the income and market approaches. In his market approach he viewed Sony/ATV as a music catalog rather than an operating music-publishing company. This meant that he calculated the venture's enterprise value by its net publishers share (NPS) and not EBITDA. Anson's income approach also differed from the Estate's in using Sony/ATV's historical financial data to predict future cashflows. Anson made no discounts based on lack of control or marketability, but did make a discount based on Sony's option to buy half of Jackson's interest in Sony/ATV. He valued NHT II at $206 million.

Anson valued Jackson's interest in Mijac using only the income approach. His major disagreement with Dahl was about the size and duration of a postdeath boom in demand for Jackson's music, as well as the number of unreleased compositions that he thought Jackson had left behind. In the end Anson valued Jackson's interest in Mijac at $114 million.

[*59] 2. Anson's Credibility

As the Commissioner's only expert witness, Anson's credibility was an especially important part of the case. And it suffered greatly at trial. His problems began when he was asked about the effect on himself and his firm if the Commissioner prevailed in the case. He responded: “I have no idea. I've never worked for the Internal Revenue Service before.” Later when asked whether he or his firm had previously been retained by the Commissioner to write an intellectual-property valuation report in Whitney Houston's estate-tax case, Anson replied: “No. Absolutely not.” That was a lie. Approximately two years before he testified, the Commissioner had retained Anson to write a valuation report titled, “Analysis of the Fair Market Value of the Intangible Property Rights Held by the Estate of Whitney E. Houston as of February 11, 2012 For Estate Tax Purposes.” It was only after a recess and advice from the Commissioner's counsel that Anson admitted to this.

Anson also testified that neither he nor his firm ever advertised to promote business. This was also a lie. In the midst of trial, Anson's firm touted his testimony in the following email blast:

What has been described as the “tax trial of the century” by the Hollywood Reporter, the case between the Internal Revenue Service and the Estate of Michael Jackson began in Tax Court this week.

[*60] CONSOR Chairman Weston Anson is the expert of the century and will be testifying on behalf of the IRS.

The big discrepancy in the value of the Jackson estate will be sure to bring testimony tailor made for a Hollywood blockbuster. While CONSOR valued the intellectual property assets of the Jackson estate at a total close to $1 billion, the estate initially valued the assets at time of death at a mere $2,105.

And in a lecture given before trial Anson referred to his valuation in this case, stating, “I'm sitting today * * * in a deposition in what's known as the 'Billion Dollar Tax Case.' * * * [W]e've just spent the last year valuing the estate of Michael Jackson.” When asked at trial whether he had in fact referred to this case as a billion-dollar case, Anson replied with his own question: “Would you like to be called the lawyer of the century?”

The Estate moved to strike all of Anson's testimony, including his expert reports, as tainted by perjury. We denied the Estate's motion finding it “too severe.” We instead stated that “[a] more proportionate remedy would be to discount the credibility and weight we give to [Anson's] opinions.” There is nothing wrong about marketing one's services or taking on another case for the IRS while working on this one. But Anson did undermine his own credibility in being so parsimonious with the truth about these things he didn't even benefit [*61] from being untruthful about, as well as in not answering questions directly throughout his testimony.

This affects our factfinding throughout.

C. Issues Left for Decision

We are left to wade through these facts to decide the fair market value at Jackson's death of the three contested assets. Depending on what we find, we must also decide whether the Estate is liable for a substantial-understatement penalty under section 6662(b)(2) or a gross-valuation misstatement penalty under section 6662(h)(2)(C).

We summarize the parties' positions:

	Reported one state return	Notice of deficiency	Estate on brief	Commissioner on brief
Image and likeness	$2,105	$434,264,000	$3,078,000	$161,307,045
New Horizon Trust II (Sony/ATV)	-0-	469,005,086	-0-	206,295,934
New Horizon Trust III (Mijac Music)	2,207,351	60,685,944	2,267,316	114,263,615

[*62] OPINION

I. Estate Tax Valuation Principles

The Code imposes a tax on “the transfer of the taxable estate of every decedent who is a citizen or resident of the United States,” sec. 2001(a), and it defines the taxable estate as “the value of the gross estate” less applicable deductions, sec. 2051. The value of the gross estate of a decedent is “the value at the time of his death of all property, real or personal, tangible or intangible, wherever situated,” to the extent provided in sections 2033 through 2045. Sec. 2031(a). Section 2033 includes in the gross estate the value of “all property to the extent of the interest therein of the decedent at the time of his death.” And the regulations tell us to value a decedent's property at its fair market value. Sec. 20.2031-1(b), Estate Tax Regs.

Fair market value is the “the price at which the property would change hands between a willing buyer and a willing seller, neither being under any compulsion to buy or to sell and both having reasonable knowledge of relevant facts.” Id. Under this standard, the hypothetical willing buyer's knowledge extends beyond just those facts publicly known, as he is “presumed to be 'reasonably informed' and 'prudent' and to have asked the hypothetical willing seller for information that is not publicly available.” Estate of Kollsman v. [*63] Commissioner, 113 T.C.M. (CCH) 1172, 1178 (2017), aff'd, 777 F. App'x 870 (9th. Cir. 2019). We must also be mindful that these hypothetical buyers and sellers are, after all, purely hypothetical, meaning we mustn't “construct[ ] particular possible purchasers” based upon “imaginary scenarios” about who they might be or how they might act. Estate of Simplot v. Commissioner, 249 F.3d 1191, 1195 (9th Cir. 2001), rev'g and remanding 112 T.C. 130 (1999); see also Morrissey v. Commissioner, 243 F.3d 1145, 1148 (9th Cir. 2001), rev'g Kaufman v. Commissioner, 77 T.C.M. (CCH) 1179 (1999); Cave Buttes, L.L.C. v. Commissioner, 147 T.C. 338, 357-58 (2016).

It is also fundamental to this approach that property is valued as if “in the decedent's hands at the time of its transfer by death.” Estate of Simplot, 249 F.3d at 1194-95 (citing sec. 2033). Because property is valued precisely at the moment of death,¹⁹ it is inappropriate, as a general matter, to ascribe value based upon postdeath evidence. The temptation to use hindsight is usually too great. See Estate of Gilford v. Commissioner, 88 T.C. 38, 52 (1987); Estate of Gallagher v. Commissioner, 101 T.C.M. (CCH) 1702, 1706 (2011). The prohibition is not absolute, however, so a court may for instance consider subsequent events “to the [*64] extent that they were reasonably foreseeable” at the decedent's death. Trust Servs. of Am., Inc. v. United States, 885 F.2d 561, 569 (9th Cir. 1989) (citing Estate of Gilford, 88 T.C. at 52). Whether a subsequent event was reasonably foreseeable is a question of relevance: Evidence of the actual price for a sale after death can be relevant “so long as the sale occurred within a reasonable time after death and no intervening events drastically changed the value of the property.” First Nat'1 Bank of Kenosha v. United States, 763 F.2d 891, 894 (7th Cir. 1985).

II. Expert Opinions

Estate-tax cases are very often disputes about valuation. And valuation disputes are questions of fact, see Estate of Gallagher, 101 T.C.M. (CCH) at 1705, that are very often battles of the experts. This is especially true here since all three assets at issue are unique and possibly of great value.

While experts are helpful, we are not bound by any particular expert opinion. Hunt & Sons, Inc. v. Commissioner, 83 T.C.M. (CCH) 1345, 1352 (2002); see also Helvering v. Nat'l Grocery Co., 304 U.S. 282, 295 (1938). We are free to accept or reject an expert's opinion based on our sound judgment. Estate of Hall v. Commissioner, 92 T.C. 312, 338 (1989). We are also free to accept only a portion of an expert's opinion. Parker v. Commissioner, 86 T.C. 547, 562 (1986); Ames v. Commissioner, 58 T.C.M. (CCH) 1470, 1488 [*65] (1990), aff'd sub nom. Hildebrand v. Commissioner, 967 F.2d 350 (9th. Cir. 1992).

III. Valuation

A. Basics

The difficulty with valuing the assets at issue here is that they are not like shares of publicly traded stock. Each of the three assets is unique, making it difficult to determine its value. There are three approaches that courts and appraisers use to value unique assets: income, market, and cost. See, e.g., Cave Buttes, 147 T.C. at 358; David Laro & Shannon P. Pratt, Business Valuation and Federal Taxes: Procedure, Law, and Perspective 162 (2d ed. 2011).

The income approach values an asset by calculating how much revenue it will produce in the future and discounting that revenue back to its present value, because a dollar today is worth more than a dollar in the future. Laro & Pratt, supra, at 163. The market approach values an asset by comparing it to the prices at which similar assets have changed hands in arm's-length transactions close in time to the date of death. Id. at 196. The cost approach values an asset by computing the cost of recreating it. See United States v. Eden Mem'l Park Ass'n, 350 F.2d 933, 935 (9th Cir. 1965); Marine v. Commissioner, 92 T.C. 958, 983 (1989), aff'd without published opinion, 921 F.2d 280 (9th Cir. 1991).

[*66] All the experts here use the income approach to value the three assets at issue, although the market approach pops up in conjunction with the income approach for one of the assets. None of the experts used the cost approach for any of the assets.

B. Discounted Cashflow Method and Its Discount Rate

The income approach has two variations: discounted cashflow (DCF) and capitalization. See Cave Buttes, 147 T.C. at 358; Laro & Pratt, supra, at 162. Both parties here use only the DCF method. The DCF method has two main variables: the projected future cashflow stream and the discount rate. The parties here largely disagree about each of the three assets' future cashflow projections, and we will address those differences below.

Projections of cashflow are distinct from the selection of an appropriate discount rate — a variable that can have a large effect on the outcome of a DCF analysis. Laro & Pratt, supra, at 164. The discount rate accounts for the time value of money in computing the present value of future cashflows an asset is projected to produce. See Shepherd v. Commissioner, 115 T.C. 376, 392 (2000), aff'd, 283 F.3d 1258 (11th Cir. 2002); Laro & Pratt, supra, at 168. The experts here largely agree on the formula, but not the inputs, that we should use to calculate the discount rate.

[*67] This formula computes a discount rate in the form of a weighted average cost of capital (WACC). The formula is:

• E = market value of equity;

• D = market value of debt;

• r_e = cost of equity; and

• r_d = cost of debt.

The variables in this formula are relatively straightforward — only the cost of equity is itself derived from a formula with more variables in it. To determine the cost of equity, each expert used the capital-asset pricing model.²⁰ The formula to calculate the cost of equity using the capital-asset pricing model is:²¹

[*68] r_e = r_f + β * (r_m − r_f)

• r_e = cost of equity;

• r_f = risk-free rate;

• β = beta; and

• r_m = expected return of the market.

The risk-free rate takes into account a riskless security — typically a U.S. government security such as a Treasury bond. AEP Tex. N. Co. v. Surface Transp. Bd., 609 F.3d 432, 436 (D.C. Cir. 2010). The beta measures the covariance between the rate of return on a company's stock and the overall market return — i.e., systematic risk.²² See Furman v. Commissioner, 75 T.C.M. (CCH) 2206, 2214-15 (1998). The market-risk premium is the difference between the expected market return over the risk-free rate. See Hoffman v. Commissioner, 81 T.C.M. (CCH) 1588, 1598-99 (2001).

[*69] While the parties agree that we should use the DCF method, they disagree as to what the discount rate should be because they disagree on the values of several of the variables in both of these formulas.

C. Synergy

Another important difference between the parties is what at trial was called synergy. This argument lurked in the background throughout the trial — Anson kept trying, especially when he discussed the value of Jackson's image and likeness, to include the value of other assets, such as Jackson's copyrights in his musical compositions and performances, in the value of the assets at issue. His justification is that these assets all belonged to the Estate and would be more valuable in some circumstances if they could be bundled and exploited together.

One can't help but notice how often Anson's valuations — both in how he chose to describe the assets that he valued and in how he valued them — seemed to keep arriving at places that the Estate did in fact come to, albeit after the assets were no longer under Jackson's control and when they had been managed with stunningly greater competence than they had been in Jackson's own hands.

From the Estate's perspective, this kind of valuation is just hindsight, not even 20/20 hindsight but more like that of an eagle or a spy satellite. But there's something more to this — a pair of problems and not just one. The first problem is [*70] how to value any asset that requires active management, because it is so difficult to distinguish between the value of that asset and the value of its management. As a purely theoretical exercise, one might imagine that the right way to value an estate with such assets is to imagine an auction at the side of the decedent's deathbed, with the auctioneer seeking a single price for all his assets. This would make it easy to determine — especially in an extreme case like Jackson's, with its very difficult-to-value assets that turned out to be quite lucrative — the total value of those assets as of the date of death. One would just run a hypothetical auction and take as the date-of-death value the second highest bid. See generally Whitehouse Hotel Ltd. P'ship. v. Commissioner, 139 T.C 304, 332-37 (2012), aff'd in part, vacated in part and remanded, 755 F.3d 236 (5th Cir. 2014). The difference between this bid and the winning one would then be the expected marginal contribution of the winning bidder in managing the estate. This would allow a hypothetical judge to easily distinguish the value of a decedent's assets from the value added by the management of those assets.

A second and distinct problem is how to measure the effect that separate assets can have on each other's value. It would be entirely reasonable to think that a collection of related intellectual-property rights, such as copyrights in music, recordings and images, might be more valuable if they could be packaged and sold [*71] together. Anson called these synergies, but they are really a problem of transaction costs. For example, when valuing 100 shares of stock in a company with only 150 shares outstanding, does one value each share separately, or does one value all 100 shares together? We have acknowledged in the past that a premium may be appropriate when valuing large blocks of stock. See Estate of Mitchell v. Commissioner, 83 T.C.M. (CCH) 1524, 1528 (2002); Estate of Salsbury v. Commissioner, 34 T.C.M. (CCH) 1441, 1451-52 (1975). And the regulations tells us that the degree of control is relevant to valuation. See sec. 20.2031-2(e), Estate Tax Regs.; sec. 25.2512-2(e) and (f), Gift Tax Regs. As stated in Revenue Ruling 59-60, sec. 4.02(g), 1959-1 C.B. 237, 242, “[t]he size of the block of stock itself is a relevant factor to be considered * * * [and] may justify a higher value for a specific block of stock.”

When a court values a block of 100 shares as worth more than 100 individual shares, it nods towards life in the real world. In a world without transaction costs, the 100 holders of 1 share each could get together and do with the corporation anything that one holder of 100 shares could do. But there is some real-world caselaw here. It focuses on the value of the nature of the estate tax as a “tax on the privilege of passing on property, not a tax on the privilege of receiving property.” Ahmanson Found. v. United States, 674 F.2d 761, 768 (9th Cir. 1981). [*72] A decedent who owns a controlling block of shares in a corporation may in his will leave them to 100 legatees, none of whom would have a controlling share. But the cases that discuss this question hold that the taxable value of such an estate includes the value of the controlling block: “There is nothing in the statutes or in the caselaw that suggests that valuation of the gross estate should take into account that the assets will come to rest in several hands rather than one.” Id.; see also Estate of Curry v. United States, 706 F.2d 1424, 1427-28 (7th Cir. 1983).

This is the unarticulated point that Anson and the Commissioner make here: Jackson's will didn't divide his valuable intellectual property; it kept it together. Because keeping it together made dealmaking much easier — all those valuable rights could be bundled, as they in fact were by Branca in the years after Jackson's death — shouldn't they be valued together, like a controlling block of stock?

Both these problems are reasonable in their statement (or maybe our restatement) of them. But we aren't in a position to conduct theoretical deathbed auctions, see Estate of Simplot, 249 F.3d at 1195, and we don't let parties out of their stipulations easily, see Stamm Int'l Corp. v. Commissioner, 90 T.C. 315, 321-22 (1988). Instead, to address these two problems, we will stick to the solutions that precedent dictates.

[*73] How do we disaggregate the value that Branca added to the Estate from the value of those assets themselves? What we will do is what we've always done — separate facts known or knowable at the date of death from those remoter in time or unforeseeable, and then ask what a hypothetical buyer in possession of these facts would offer for them in an arm's-length deal with a similarly knowledgeable hypothetical seller. And, we reiterate, we do not hypothesize a particular buyer or a particular seller with any particular skills or use for those assets. See Estate of Giustina v. Commissioner, 586 F. App'x 417, 418-19 (9th Cir. 2014), rev'g and remanding T.C. Memo. 2011-141; Estate of Simplot, 249 F.3d at 1195.

How do we figure out whether to value several assets individually or as a block? We look to how the parties prepared the case. Form 706, United States Estate (and Generation-Skipping Transfer) Tax Return, tells executors to list miscellaneous assets in an estate (and that's what intellectual-property assets are). See 2009 Form 706, at 3; Instructions for Form 706, at 26. That's what the Estate did here — listing several intangible assets in Schedule F, Other Miscellaneous Property — including:

• image and likeness;

• MJJ Productions, Inc.; and [*74]

• master recordings.

The Estate separately reported Jackson's interests in NHT II and NHT III on Schedule G, Transfers During Decedent's Life. And it reported his interest in MJJ Ventures, Inc., on Schedule B, Stocks and Bonds.

The Commissioner in his notice of deficiency described his disagreements with the values that the Estate reported, but he did not object to this list of what those assets were. The parties spent years in discovery and other pretrial preparation, at the end of which they reached stipulations about some of these assets (such as master recordings, MJJ Productions, and MJJ Ventures). The Commissioner could have chosen to object to the description of the assets to be valued, or he could have refused to stipulate the values of all these assets rather than agree to some and not agree to others. But what he's not allowed to do is renege on his stipulation to cram the value of assets whose value he stipulated into the value of assets whose value he did not stipulate. That would undermine the stipulation, which we don't allow the parties to do absent mutual mistake or proof that a party was misled, or if justice requires it. See Rule 91(e); Stamm Int'l Corp., 90 T.C. at 321-22; Buchsbaum v. Commissioner, 83 T.C.M. (CCH) 1777, 1779 (2002).

[*75] The result of the way this case came to trial means that we will value only those three assets whose values the parties couldn't agree about. In making findings on those values we must be clear on what we consider. We will assume that the Estate's assets can be used together to generate value. But we will not add the value of other assets to the value of the assets at issue.

IV. Tax Affecting²³

A. The Basics

A second consideration that affects our valuation of all three assets here is tax affecting. Each of the Estate's experts takes tax affecting into account in his valuations. To understand tax affecting and why its proponents claim it is necessary, we begin with a short tax-vocabulary lesson.

Each asset in dispute in this case is held by a pass-through entity, which means the Code imposes no tax on the income that these assets produce.²⁴ Such entities are not at all exotic — they include common forms of private business [*76] ownership such as partnerships, S corporations, and LLCs. When one of these entities earns income, it passes right through to its partners, shareholders, or members who themselves have to pay tax on the income at their individual rates. This is in contrast to income from C corporations.²⁵ C corporations don't pass income through to their shareholders; they get taxed on it themselves. And then the Code taxes any income that trickles through to their shareholders as dividends at their individual rates. This two-layer tax on C corporation income is known as “double taxation.” See, e.g., Pierre v. Commissioner, 133 T.C. 24, 30 (2009), supplemented by T.C. Memo. 2010-106. Because of this, valuation experts refer to the entity-level income of pass-throughs as pretax, and C corporations' income as after-tax. Investor-level income — the income that ultimately ends up in an investor's pocket — will always be after-tax.

The choice of entity can have big effects on the tax consequences that a business faces. For example, in 2009 the top marginal tax rate on C corporation income was 35%, sec. 11(b); dividends were typically taxed at 15%; sec. 1(h)(1), (11); and the top marginal rate for individuals was 39.6%, sec. 1(a). If the sole [*77] asset of a pass-through entity and the sole asset of a C corporation each produce $100 in income, entity-level income of the pass-through and the C corporation will be $100 (= $100 * (1 − 0.0)) and $65 (= $100 * (1 − 0.35)), respectively. The pass-through and C corporation's investor-level income, however, will be $60.40 (= $100 * (1 − 0.396)) and $55 (= $65 * (1 − 0.15)).²⁶

Other things being equal, an income-producing asset would thus be worth more to a pass-through than to a C corporation. But the complications of the real world intrude. Even with this tax disadvantage, C corporations dominate the financial markets and are by far the largest share of the largest business organizations in America. See generally James D. Cox & Thomas Lee Hazen, Business Organizations Law secs. 1.5-.6 (4th ed. 2016) (describing the advantages and disadvantages of corporations). And that creates a problem for a DCF analysis. Because pass-throughs do not pay tax at the entity level, their projected cashflows will not account for any tax consequences. But almost always — and, in fact, in this case — the rate used to discount projected cashflows to present value is derived from after-tax, publicly available C-corporation information. Proponents of tax affecting argue that this mismatch between pretax cashflows and after-tax [*78] discount rates must be corrected, or tax affected. We have acknowledged as much in our cases, stating: “[I]f, in determining the present value of any future payment, the discount rate is assumed to be an after-shareholder-tax rate of return, then the cash-flow should be reduced ('tax affected') to an after-shareholder-tax amount. If, on the other hand, a preshareholder-tax discount rate is applied, no adjustment for taxes should be made to the cash-flow.” Gross v. Commissioner, 78 T.C.M. (CCH) 201, 209 (1999), aff'd, 272 F.3d 333 (6th Cir. 2001). Whether, and precisely how, to tax affect a pass-through's earnings, however, is the subject of significant dispute in this case.

B. The Experts' Positions

Fishman, Wallis, and Dahl in their respective DCF analyses concluded that the appropriate hypothetical buyer for each asset would be a C corporation, and therefore, each of them reduced cashflows by the income-tax liability that would be paid by a hypothetical C corporation buyer. To make things even more complicated, each also computed a discount rate that included the effects of a C corporation's tax rate. They all stated that this was appropriate because it used both after-tax cashflows and after-tax discount rates. Each of the Estate's experts, however, used a different tax rate to do his computation:

• Fishman applied a 35% rate based on the federal rate, [*79]

• Wallis applied a 39.615% rate based on a combined federal and New York State rate, and

• Dahl applied a 39.8% rate based on a combined federal and undisclosed state rate.

When we've faced this issue in the past, we've shied away from tax affecting because of these difficult practical problems. See Estate of Gallagher, 101 T.C.M. (CCH) at 1710; Estate of Giustina, 101 T.C.M. (CCH) at 1679; Dallas v. Commissioner, 92 T.C.M. (CCH) 313, 317-18 (2006) (tax affecting not appropriate when the taxpayer presumed that an S corporation would lose its S corporation status after a sale); Gross, 78 T.C.M. (CCH) at 207. For example, in Wall v. Commissioner, 81 T.C.M. (CCH) 1425, 1432-33 n.19 (2001), we noted that

[t]he argument in favor of tax-effecting stresses that many potential buyers of S corporations are C corporations. Because a C corporation would be unable to maintain a target company's S corporation status following an acquisition, the C corporation would tax-effect the S corporation's income (at C corporation rates) in deciding how much it would pay for the S corporation. See Trugman, Understanding Business Valuation: A Practical Guide to Valuing Small to Medium-Sized Businesses, at 198-199 (1998). By contrast, the argument against tax-effecting stresses that although an S corporation's stockholders are subject to tax on the corporation's income, they are generally not subject to a second level of tax when that income is distributed to them. This could make an S corporation at least somewhat more valuable than an equivalent C corporation. However, tax-effecting an S corporation's income, and then determining the [*80] value of that income by reference to the rates of return on taxable investments, means that an appraisal will give no value to S corporation status.

There has, it seems, been only one case where we allowed tax affecting in a valuation. See Estate of Jones v. Commissioner, T.C. Memo. 2019-101, at *41-*42. In Estate of Jones, both experts agreed that a hypothetical buyer and seller would take into account the form of business entity in determining the fair market value of a limited-partnership interest. Id. at *39. The parties just disagreed on how to account for this effect. Id. The Commissioner's expert argued against tax affecting because the company at issue was a natural-resource holding company — not because it would pay no entity-level tax. Id.

The experts here strongly disagree on the appropriateness of tax affecting. We view this disagreement just as we have in the past, as one that is a dispute about fact. And we find, as we have done consistently in the past apart from Estate of Jones, that by a preponderance of the evidence tax affecting is not appropriate here because the Estate has failed to persuade us that a C corporation would be the hypothetical buyer of any of the three contested assets. The Estate's experts did not even discuss in a persuasive way their reasons for assuming that a C corporation would be the only or even likely buyer for these assets. Fishman, for example, reasonably points out that any buyer of Jackson's likeness and image [*81] would have to spend significant amounts of money to rehabilitate and defend its value. He claims a C corporation would be the hypothetical buyer of Jackson's image and likeness because history shows us that C corporations have bought the image and likeness and associated trademarks of other celebrities.

It is possible that rehabilitating an image as tattered as Jackson's had become would require capital outlays more typical of public corporations. But we don't think that's more likely than not to be true. Fishman valued Jackson's image and likeness, which if the Estate is to be believed is the most valuable of the three contested assets, at just over $3 million. That is not a sum so large as to make it likely that only a C corporation would be able to buy it. There has also been a boom in different types of pass-through entities — such as limited liability companies — that give organizations the many benefits of C corporations in raising capital from large numbers of shareholders or members while avoiding double taxation. Cox & Hazen, supra, secs. 1.1, .7(6), .9, .11. Many of our precedents arose from S corporations, which have sharp restrictions on who and what can own them. With the advent and popularity of other, less restrictive, forms of pass-through ownership we cannot but find that the gap between C corporations and other entities has narrowed over time. The Estate's experts did not consider [*82] such distinctions and did not consider both the tax detriments and benefits of pass-through status.

Our finding reflects these facts: The Estate's own experts used inconsistent tax rates. They failed to explain persuasively the assumption that a C corporation would be the buyer of the assets at issue. They failed to persuasively explain why many of the new pass-through entities that have arisen recently wouldn't be suitable purchasers. And they were met with expert testimony from the Commissioner's side that was, at least on this very particular point, persuasive in light of our precedent. This all leads us to find that tax affecting is inappropriate on the specific facts of this case. We distinguish Estate of Jones as an instance where the experts agreed to take into account the form of the business entity and agreed on the entity type. The Commissioner argued there, as he does here, that we shouldn't tax affect, but his own experts didn't seem to be on board. As we observed, “[t]hey do not offer any defense of respondent's proposed zero tax rate. Thus, we do not have a fight between valuation experts but a fight between lawyers.” Estate of Jones, at *39.

We do not hold that tax affecting is never called for. But our cases show how difficult a factual issue it is to demonstrate even a reasonable approximation [*83] of what that effect would be. In Estate of Jones, there was expert evidence on only one side of the question, and that made a difference.

That was not the case here.

V. Rights in Music Intellectual Property for . . . Tax Lawyers

Jackson's most valuable assets were intellectual property in his image, in his own music, in others' music that he bought when he was at the peak of his popularity, or rights indirectly dependent on those rights. This is a specialized area of law whose terms we've already used in our factfinding. Before we plow into the complex valuation analysis that lies ahead, any tax specialists reading this might benefit from a primer on three key concepts: composer, performer, and right of publicity.

A. Composer

In copyright law, a composer is someone who writes compositions. A composition is the combination of words and music that is performed. Music, however, is often a collaborative trade, and a composition's lyricist and composer are often two different people. This is no surprise to fans of musical theater who are familiar with Gilbert & Sullivan, Lerner & Loewe, Rodgers & Hammerstein, and Sondheim and himself. But in popular music there seems to be less specialization. There are singer-songwriters, singers of original compositions, and [*84] cover artists — those singers who perform songs already sung by others. A concrete example that is relevant here is Temperton, who is the composer of “Thriller” — he didn't actually perform the song, but he wrote the lyrics and created the tune.

A composer gets rights to his composition under copyright law. “[C]opyright ownership in the musical composition initially vests in its creators.” 6 Melville B. Nimmer & David Nimmer, Nimmer on Copyright, sec. 30.02 (Mathew Bender 2020). Songwriters of the two elements of a composition — music and lyrics — each acquire by law an equal right to the copyright in that musical composition. See Greene v. Ablon, 794 F.3d 133, 151 (1st Cir. 2015); Garcia v. Google, Inc., 786 F.3d 733, 751 n.1 (9th Cir. 2015) (Kozinski, J., dissenting from en banc ruling); Richlin v. Metro-Goldwyn-Mayer Pictures, Inc., 531 F.3d 962, 968 (9th Cir. 2008); Nimmer, supra, sec. 30.02[A]. See generally Maurel v. Smith, 271 F. 211, 215-16 (2d Cir. 1921); Cal. Civ. Code sec. 981(a) (West 1982) (common-law copyright). A composer's copyright attaches to his work automatically, though songwriters will often submit it to the U.S. Copyright Office²⁷ to ensure protection.²⁸

[*85] 1. Income Streams

A major factor in the complexity of the competing valuations in this case is that composition copyrights in music are the fount of several income streams and not one big river. There are four that are important here:

• mechanical royalties,

• performance royalties,

• synchronization fees, and

• other miscellaneous income.

Mechanical royalties were originally income from sales of a physical object such as a vinyl record or a CD. Even as the disintermediation of albums into infinitely reproducible digital downloads of individual songs began, the royalties collected by composers for each of their downloaded compositions continued to be called mechanical royalties. Mechanical royalties are the income to composers [*86] from the sales of those physical or digital embodiments of a recorded performance.

Performance royalties are the income composers get when their compositions are performed live, on the radio, or on television. This income flows from live concerts where the song is performed, from radio or television broadcast where a recorded song is played, or from a restaurant or club when a song is performed or a recorded version is played.

One's thoughts should naturally turn again to transaction costs. The right of a composer to performance royalties may extend to hundreds of millions or billions of televisions and radios, and it'd be impossible to monitor all the individual performances. As recognition of copyright in music spread, the solution was to create performing-rights organizations. Composers contract with an organization like BMI or ASCAP (the American Society of Composers, Authors, and Publishers). These organizations create and sell blanket licenses to those who want to play copyrighted compositions, and the resulting royalties are divided among the copyright holders. The organization makes this division by extensive sampling on what's being played on the radio or television or as background in bars and restaurants.

[*87] Synchronization income — better known as a “sync fees” — is “paid whenever a piece of music, in whole or part, is synchronized in a timed relation to the visual image.” This stream generally comes from advertising, film, television, and video games. Think of a movie's soundtrack or an advertising tune that one hears on TV.

Other miscellaneous income is minor. There is still a small market, for example, for sheet music or song folios.

2. Publishers

A composer who wants upfront money for these usually uncertain income streams, or one who wants someone else to manage the negotiations for use of his music, or the young and naive talent who signs away his songs for a song, all have to deal with the music-publishing industry. Publishing has an industry-specific meaning here. Publishers, e.g., Mijac, are buyers or assignees of the copyrights to compositions. Publishers gain “[t]he power to control the exploitation of a musical composition, and an interest in the income generated from such exploitation.” Nimmer, supra, sec. 30.02. In exchange, they usually agree to provide composers services such as copyright registration, promotion, and a mechanism to collect income throughout the world; and to pay an advance. A publisher typically has better access to the industries that want to buy the right to [*88] use a composition. Movie studios, advertising companies, and the like find it more efficient to browse through a large publisher's catalog rather than deal with often numerous composers of a single song. Though the publisher usually owns the copyright to a composer's work, the royalties earned are split between the publisher and the composer.

How are the royalties split? Bringing a familiar tax concept into the world of music — it depends. The contracts between composers and publishing companies vary. The split can be a percentage of revenue generated or a fixed fee. The publisher may ask for rights that equal the life of the copyright (which is typical) but the time may be shorter. Sometimes composers will simply pay the publisher to administer the songs. Such an administration contract limits the publisher's role to the collection of income and some promotional work, and usually generates less income for the publisher.

But a publisher can't collect this revenue forever. Songs written before 1978 are subject to the Copyright Act of 1909. See Penguin Grp. (USA) Inc. v. Steinbeck, 537 F.3d 193, 197 (2d Cir. 2008); Self-Realization Fellowship Church v. Ananda Church, 206 F.3d 1322, 1325 (9th Cir. 2000). Under that Act, the duration of a copyright was 28 years and it could be renewed for another 28 years by the composer — not the publisher. Penguin Grp., 537 F.3d at 197. However, [*89] because of several amendments, these works can again be renewed for 19 years, and then again for 20 years. See Nimmer, supra, sec. 9.11[B][1]. See generally Penguin Grp., 537 F.3d at 197.

The structure is a bit different for songs written during or after 1978. Under the current Copyright Act, the duration of a copyright is the life of the author plus 70 years. 17 U.S.C. sec. 302 (1998); see also Nimmer, supra, sec. 9.10[A][1]. The author has recapture rights that allow the termination of a license to a third party beginning 35 years after the date of grant.²⁹ 17 U.S.C. sec. 203(a)(3) (2002); see also TD Bank N.A. v. Hill, 928 F.3d 259, 273 (3d Cir. 2019).

B. Performer

Jackson was even more famous as a performer than he was as a composer. Copyright law also protects performers. Who is a performer in copyright law? In its most basic sense, it's the person (or group) who performs a composition that is recorded on a tangible medium. Think of Jackson performing “Thriller”. The tangible medium for a performer is the “master recording” — “the actual recorded performance of a particular musical composition.” Allen Bargfrede, Music Law in [*90] the Digital Age 15 (2d ed. 2017). The recording artist gets rights in this master recording.

A real-world example — albeit one with tweaked facts to make it simple — might help clarify who gets what rights. In 1973 Bob Dylan recorded a song that fans called “Rock Me, Mama” — a “song” that really was just a chorus made up of lyrics he wrote and a tune he created. Shortly after, Ketch Secor, a member of Old Crow Medicine Show, used the chorus created by Dylan and wrote the verses for a song that eventually became “Wagon Wheel.” About forty years later, Darius Rucker released his own performance of that song. Darius Rucker was performing Bob Dylan's and Ketch Secor's composition. So what rights does everyone have?

Bob Dylan gets composition rights for both versions of “Wagon Wheel,” as his composition — the lyrics and music — are the ones being performed. Ketch Secor also gets composition rights to both versions of the song “Wagon Wheel,” as he wrote verses to the song. Old Crow Medicine Show — as recording artist — gets performance rights in its master recording of “Wagon Wheel.” Darius Rucker — as a recording artist — gets performance rights in his master recording of “Wagon Wheel.”

Performers, like composers, can use their copyrights to produce different types of income. The first and obvious source of revenue is concert tickets from [*91] live performances. A superstar artist can be paid a guaranteed amount or a percentage of ticket sales. His additional rights, however, are quite similar to a composer's. Performers can get mechanical royalties — a piece of the sale from every record, CD, or digital download. They can also get synch fees if their performance ends up in a commercial or soundtrack.

But one major difference is that a performer has no rights to royalties for public performance. See, e.g., Bonneville Int'l Corp. v. Peters, 347 F.3d 485, 487 (3d Cir. 2003). This means that a performer is not paid when his song is played on the radio. See 17 U.S.C. sec. 114(d)(1)(B)(iii); Bonneville, 347 F.3d at 487; Bargfrede, supra, at 19. An exception to this rule is the public performance for sound recordings in digital transmission. Digital Performance Right in Sound Recordings Act of 1995, Pub. L. No. 104-39, sec. 3, 109 Stat. at 336-44; see also Bonneville, 347 F.3d at 488-89; Bargfrede, supra, at 21. This allows a performer to collect royalties for his work if it is played through digital means such as satellite radio or a streaming service. See Digital Performance Right in Sound Recordings Act of 1995, sec. 3; Bonneville, 347 F.3d at 488-89; Bargfrede, supra, at 21.

[*92] C. Right of Publicity

The third right in music IP, and one that is very important in this case, is the right of publicity (ROP). It's a right that's not peculiar to musicians, but to celebrities generally. ROP is the right to control the commercial use of one's identity. 1 J. Thomas McCarthy & Roger E. Schechter, The Rights of Publicity and Privacy, sec. 1:3 (2d ed. 2020). It is a creation of state law.³⁰ C.B.C. Distrib. & Mktg., Inc. v. Major League Baseball Advanced Media, L.P., 505 F.3d 818, 822 (8th Cir. 2007) (citing Zacchini v. Scripps-Howard Broad. Co., 433 U.S. 562, 566 (1977)); McCarthy & Schechter, supra, sec. 1:3. ROP generally consists of two separate components: ROP and associated trademarks. One's ROP is a distinct legal category, “not just a 'kind of' trademark, copyright, false advertising or right of privacy.” McCarthy & Schechter, supra, sec 1:3. ROP protects someone's name, likeness, voice, signature, or photograph. The other component — associated trademarks — are trademarks with respect to the ROP.

ROP originated in a right sometimes recognized at common law as the right of privacy. But the right of privacy was specifically a “right to be left alone.” Id. sec. 1:25. Some celebrities don't want to be left alone, but do want to have a [*93] say in how their image is used and who uses it. When celebrities began to bring these kinds of cases under privacy statutes or the common law, courts struggled with the “right of privacy” label. Id. sec. 1:7. These celebrities didn't really want privacy in the traditional sense of being left alone, and many of these cases were therefore dismissed. Id. But if this was just a problem of nomenclature, it went away in Judge Jerome Frank's Haelan Labs., Inc. v. Topps Chewing Gum, Inc., 202 F.2d 866 (2d Cir. 1953), where he coined the phrase “right of publicity.” Id. at 868; see also McCarthy & Schechter, supra, sec. 1:26. This is the label that has stuck.

Because Jackson was domiciled in California at the time of his death, his legal interest and rights in property are determined under California law.³¹ See [*94] Morgan v. Commissioner, 309 U.S. 78, 79-80 (1940) (law of decedent's domicile creates legal interests and rights for estate tax); Milton H. Greene Archives, Inc. v. Marilyn Monroe LLC, 692 F.3d 983, 986 (9th Cir. 2012) (decedent's domicile defines ROP); Cairns v. Franklin Mint Co., 292 F.3d 1139, 1149 (9th Cir. 2002) (same). By 1972, California began to codify the ROP. McCarthy & Schechter, supra, sec. 1:4. In 1979, California supplemented this statutory right with a common-law ROP in the landmark case Lugosi v. Universal Pictures, 603 P.2d 425, 428 n.6 (Cal. 1979).³² The case was brought by the surviving spouse and the son of Bela Lugosi, the original American Dracula. See id. at 427. The court recognized a common-law ROP; but then held that it did not survive death. Id. at 428-30. The California legislature stepped in and created a [*95] statutory ROP that was descendible to heirs and assignees.³³ Comedy III Prods., Inc. v. Gary Saderup, Inc., 21 P.3d 797, 799-800 (Cal. 2001). It is now codified in California Civil Code section 3344.1.³⁴ Since the common-law right still doesn't survive death, see Comedy III Prods., 21 P.3d at 799; 6A Romualdo P. Eclavea et al., California Jurisprudence 3d, Assault and Other Willful Torts, sec. 165 (West 2021), it has little effect on this case.

The dispute between the Estate and the Commissioner about the value of Jackson's ROP is in part a dispute about the value of the Estate's rights under this California statute. We need to give it a closer look: Section 3344.1(a)(1) of the California Civil Code states that

[a]ny person who uses a deceased personality's name, voice, signature, photograph, or likeness, in any manner, on or in products, merchandise, or goods, or for purposes of advertising or selling, or soliciting purchases of, products, merchandise, goods, or services, without prior consent from the person or persons [who have been given the right of consent], shall be liable for any damages sustained by the person or persons injured as a result thereof.

[*96] The boundaries of what is protected are at times fuzzy, and the situation is complicated by the survival of California's evolving common-law ROP. Caselaw helps. We'll begin with what “likeness” means. In Midler v. Ford Motor Co., 849 F.2d 460, 463 (9th Cir. 1988), the Court held “likeness” means only visual images under the statute. This might seem clear, but then someone came up with the idea of a robot version of Vanna White, the celebrity tile turner. The Ninth Circuit held in White v. Samsung Elecs. Am., Inc., 971 F.2d 1395, 1397 (9th Cir. 1992), that a robot that resembled White was not her “likeness” under the statute but left open the possibility that a sufficiently detailed manikin might be. On the other hand, in Int-Elect Eng'g, Inc. v. Clinton Harley Corp., No. C-92-20718 JW, 1993 WL 557639, at *4 (N.D. Cal. June 24, 1993), the court held that “likeness” can mean any unique and readily identifiable artistic creation.

“Voice” under the statute only covers the use of the claimant's actual voice. See White, 971 F.2d at 1397; Midler, 849 F.2d at 463. “Photograph” is better defined. California Civil Code section 3344.1(i) states that a photograph is any still or moving film where a person is readily identifiable. That same subsection defines “readily identifiable” as whether someone viewing the photograph with the naked eye “can reasonably determine who the person depicted in the photograph is.” Id.

[*97] While it is important to clarify what is protected, it is just as important — and maybe more important to this case — to point out what is not protected. “[A] play, book, magazine, newspaper, musical composition, audiovisual work, radio or television program, single and original work of art, work of political or newsworthy value, or an advertisement or commercial announcements for any of these works” is not protected under section 3344.1(a)(1). Id. sec. 3344.1(a)(2); see also Comedy III Prods., Inc., 21 P.3d at 800. By its plain language, musical compositions — of particular interest here — are not protected. See Cal. Civ. Code sec. 3344.1(a)(2). And Laws v. Sony Music Entm't, Inc., 448 F.3d 1134, 1139-43 (9th Cir. 2006), established that master recordings are not protected because they are governed by the federal Copyright Act.

As Laws suggests, there is friction when this state-created right rubs up against federal law. A person's name, for example, can be trademarked. Though a name is also protected under ROP, these two rights are not identical. McCarthy & Schechter, supra, sec. 6:141. And federal law can preempt state law. Laws, 448 F.3d at 1141. In short, federal law — though at times it may overlap with ROP — is a distinct source of separate rights. The fuzziness about the boundaries of California's ROP is an important part of the disagreement between the parties that we will analyze at some length below. See infra pp. 133-48.

[*98] One last matter of housekeeping is needed here. Throughout the trial, the parties and their experts referred to this intangible asset primarily as “image and likeness,” but also as “name and likeness,” and ROP. All the parties do seem to be valuing the same thing — Jackson's ROP rights under the California statute that we just limned. We will somewhat arbitrarily choose to call this asset “image and likeness” for the remainder of the opinion — following the lead of the parties and their experts.

We can now turn to the core of this opinion — the valuation of the three assets at issue.

VI. Image and Likeness

A. The Code

The first asset that we'll analyze is Jackson's image and likeness as defined under California Civil Code section 3344.1. And the first question we have to ask about this asset is whether it should be included in his gross estate at all. That might seem trivial here in a case where both parties spent a goodly sum on expert witnesses to appraise just how valuable a right it is. But it is a nontrivial question — remember that this right's origin is in a common-law right of privacy. McCarthy & Schechter, supra, sec. 1:4. One might ponder that if image and likeness is included in a person's gross estate, would it not practically require those heirs to [*99] exploit the image and likeness of the dead — whether or not that was the decedent's wish. Might this — again as a practical matter — compel the invasion of privacy? This is something of a novel issue for us as we haven't had a case directly addressing the taxability of the image and likeness. (And we have found only one case on point anywhere else, Estate of Andrews v. United States, 850 F. Supp. 1279 (E.D. Va. 1994).)

But we don't need to look to caselaw to answer this question. The plain language of the Code is enough. Section 2051 defines the taxable value of an estate as the value of the gross estate less deductions. The value of the gross estate includes “the value at the time of * * * death of all property, real or personal, tangible or intangible, wherever situated.” Sec. 2031(a). Since Jackson's image and likeness is an intangible right that transfers after death, see Cal. Civ. Code sec. 3344.1; Comedy III Prods., Inc., 21 P.3d at 800, it must be included in his gross estate.

The default rule is that the value of a right is its fair market value. Sec. 20.2031-1(b), Estate Tax Regs.; Elkins v. Commissioner, 140 T.C. 86, 114-15 (2013), aff'd in part, rev'd in part, 767 F.3d 443 (5th Cir. 2014). The fair market value is generally its value in its highest and best use on the valuation date. See Estate of Kahn v. Commissioner, 125 T.C. 227, 240 (2005); Estate of Mitchell v. [*100] Commissioner, 101 T.C.M. (CCH) 1435, 1438-39 (2011). The Code provides some exceptions. See, e.g., sec. 2032A (exception for family farms); Van Alen v. Commissioner, 106 T.C.M. (CCH) 427, 430 (2013). But it makes these exceptions explicit. And here, there are none that apply. We must include the value of Jackson's image and likeness at its highest and best use.³⁵

B. Summary of the Parties' Positions

1. The Estate

i. On the Return

The novelty of valuing image and likeness led to a very large difference of opinion between the parties. The Estate reported Jackson's image and likeness on its return as worth only $2,105. This might seem absurd when one recalls Jackson's fame, but the Estate's position was based on an appraisal from Moss [*101] Adams, a large and reputable accounting firm. And Moss Adams did focus entirely on the value of the Estate's opportunity to license merchandise with Jackson's image and likeness. It consciously abstained from valuing the cashflow from Jackson's copyrights in his music, as both a performer and a composer.

This appraisal was based on royalty statements, contracts, financial statements, and business plans. What Moss Adams discovered was that in the years before Jackson died and when he was in dire need of income, he had earned close to nothing from his image and likeness. This cannot be a surprise — allegations that a celebrity molested little boys might reasonably be thought to repel potential licensees in any society that has not become completely decadent. Those allegations had a dramatic effect on Jackson's ability to win sponsorships and merchandising deals once they became public. The fact that he earned not a penny from his image and likeness in 2006, 2007, or 2008 shows the effect those allegations had, and continued to have, until his death.

Moss Adams did not rely only on this historical financial data. It also looked at Jackson's scores on the Davie-Brown index — a quantitative measure of a celebrity's reputation derived from public surveys. Jackson's Davie-Brown scores showed just what one might expect — that he was one of the most recognized celebrities in the world but one who ranked among the lowest in trustworthiness as [*102] well as other important characteristics that corporations consider when looking to use someone's image and likeness to promote their products. Taking all this information into account, Moss Adams — using a DCF analysis — determined the value of Jackson's image and likeness was $2,105.

With this valuation in hand, Moss Adams went to the Estate. The Estate was surprised. Moss Adams was, after all, valuing the image and likeness of one of the best known celebrities in the world — the King of Pop — at the price of a heavily used 20-year-old Honda Civic.³⁶ Moss Adams nevertheless gave this valuation to the Estate to rely on for the estate-tax return.

When asked if he stood by his opinion at trial — over seven years after the report — Dahl (a former principal of Moss Adams) responded that he “absolutely does.” He emphasized that though Jackson's image and likeness may have increased in value after he died, all information available at the time of his death showed that there was “a deeply flawed person with serious significant issues when it came to the issue of licensing image and likeness in a non-music setting.”

In the end, the Estate deemed this valuation credible and relied on it while filling out its return.

[*103] ii. At Trial

The value of a person's image and likeness may not depend entirely on his reputation for the few years before his death. It is a right that his heirs will have for decades to come, and there is some nonzero chance that Antony was speaking antiphrastically when he said “the evil that men do lives after them; the good is oft interred with their bones.” William Shakespeare, The Tragedy of Julius Caesar act 3, sc. 2, ll. 83-84 (Alvin Kernan ed., Yale Univ. Press rev. ed. 1959). The Estate seems to have thought so — in the course of preparing for trial the Estate brought in two more experts to value Jackson's image and likeness. The main expert was Jay Fishman — managing director of Financial Research Associates. Fishman valued Jackson's image and likeness at just over $3 million.

Fishman began by estimating the revenue that Jackson's image and likeness would generate over its legal life of 70 years. Fishman credibly testified that his starting point in valuing an asset like this is usually management projections of revenue, but that Jackson's management had no such projections.

Without those, Fishman had to rely instead on historical data. His problem, as we've already found, was that Jackson had no material earnings from his image and likeness for several years before his death. This caused Fishman to be hesitant in relying solely on recent historical data because it led to a nominal value.

[*104] He instead sought help from Mark Roesler — chairman and CEO of Celebrity Valuations, LLC & CMG Worldwide, Inc. Roesler's expert report provided image-and-likeness revenue projections for 10 years following Jackson's death, as well as other background information.

Roesler considered three main factors in determining a 10-year projection of revenue:

• predeath revenue,

• postdeath rights, and

• growth and decline rates using predeath marketability and a potential postdeath boom.

Predeath Revenues. Roesler's projections took into account Jackson's image and likeness as well as associated trademarks. These were his rights under California law, any common-law and federal trademarks Jackson registered before he died, and common-law trademarks that survived his death. These trademarks included 35 foreign trademark registrations of his name and 7 of his signature and 2 U.S. federally registered marks: “MICHAEL JACKSON” and a design that depicts his famous ability to dance on his toes.

But first Roesler had to figure out whether there would be a source of future income. A celebrity is generally able to exploit his popularity in two ways: [*105] personal service and the licensing of his image and likeness. Personal-service income — such as Jackson's agreements to appear in Pepsi commercials — ends with death. Roesler therefore was left to appraise only whatever image-and-likeness revenue the Estate could anticipate.

He couldn't rely only on Jackson's historical image-and-likeness revenue. Deals to exploit a living celebrity's image and likeness are very often hitched to promises to perform personal services. Roesler came up with ten factors to divide revenue from a deal attributable to personal service from revenue attributable to image and likeness. His factors are:

• the licensee's history,

• the budget for prospective use,

• type/nature of use,

• role of celebrity in the campaign,

• geographical scope (thereby population exposure),

• type of media used,

• length of campaign,

• exclusivity,

• other intellectual-property rights held by third parties or entities, and

• synergy with the company/product.

[*106] Roesler focused on the use of Jackson's image-and-likeness rights to promote a specific product, entity, brand, or tour. He didn't consider any one-off film, TV, or personal appearances or performances; magazine shoots or interviews; autobiography authorships; residuals; or live concert performances that lacked an associated merchandising deal. He discovered that nearly all Jackson's agreements for use of his image and likeness during his life were inextricably entwined with personal services, most of them with a very significant amount of personal service required.

Roesler looked through 30 years of data to compute Jackson's historical image-and-likeness revenue.³⁷ Over the course of those 30 years, Jackson produced average annual image-and-likeness revenue of more than $2.7 million — but a significant portion of this was connected in some way to his concert tours. Roesler decided to calculate a weighted average — with more weight on deals that Jackson struck later in his career.³⁸ His weighted average annual revenue came to about $1.6 million. He then looked at other celebrities' [*107] comparable postdeath data to supplement what he'd found. He chose Marilyn Monroe, James Dean, Bettie Page, Jackie Robinson, Princess Diana, and Elvis Presley as comparable celebrities. These celebrities' adjusted image-and-likeness revenue ranged from $52,000 (Princess Diana) to $3.1 million (Marilyn Monroe). This placed his estimate of Jackson's earnings in the middle of the range. He did not believe Jackson could realistically command anything more due to his poor personal reputation.

Postdeath Image-and-Likeness Rights. Roesler was also careful to exclude Jackson's musical works from his analysis of the value of Jackson's image-and-likeness rights — although he assumed that there would be full cooperation from the various music rights holders for licenses at reasonable rates so that all music rights on the date of Jackson's death would be available for licensing.

Growth and Decline Rates. Roesler next estimated how this revenue stream would surge and ebb as time passed. He began with Jackson's marketability over the same 30-year period that he analyzed for image-and-likeness revenue. He looked at five “marketability factors:”

• reputation/appeal,

• Q scores, [*108]

• other intellectual-property rights holders/market limitations,

• past comparable deals and agreements, and

• synergy with a company's product.

Of these five factors, Roesler believed the first factor — a celebrity's reputation and general appeal — was the most important. Revenue is greatest when a celebrity is associated with positive qualities or has a good reputation. Bad conduct doesn't disable a celebrity from exploiting his image and likeness, but it does mean some money will have to be spent rehabilitating it. Roesler listed more factors that he looked at to see whether rehabilitation would even be possible:

• nature of the bad conduct,

• extent of public media exposure to the bad conduct,

• the celebrity's prescandal image and reputation,

• the celebrity's behavior after a scandal surfaces, especially acts of atonement,

• the outcome of the scandal over time, and

• whether the celebrity had died.

Though we won't go into detail for each factor, a few are worth mentioning. While there is a range of bad conduct, Roesler concluded — and we specifically find him entirely reasonable on this point — that conduct exhibiting alleged moral [*109] turpitude can have long-term effects and may well lead to public rejection. Rehabilitation will largely depend on whether the bad conduct shocks the public's moral conscience. He reasoned that heavier media exposure of bad conduct makes it harder to rehabilitate an image and that how a scandal was resolved makes a difference as well. Appearing on a late-night talk show to be asked “What the hell were you thinking?” and answering “I did a bad thing” may work for a low-grade one-time offense. Time and the public's apparent willingness to forgive and forget may soothe a scandal's effect. Even death may help by making the public more willing to forgive — or at least shift its focus to a celebrity's positives. And Roesler thought that the public never speaks unanimously — there will almost always be people who recognize a celebrity for his professional talent and skill, despite squalid personal behavior.

A witness as numerate as Roesler wanted data, and he found it in the form of Q scores — the second factor. Q scores serve as the industry standard for measuring the consumer appeal of personalities and licensed properties — or, more simply put, how likely the celebrity can be a corporate spokesman. A positive Q score is the ratio of the number of people surveyed who chose the individual as “one of their favorites” to the number who state they are familiar with the individual. A negative Q score is the ratio of the number of people who rate an [*110] individual as “fair/poor” to the number of people familiar with the individual.

The third factor — other intellectual-property rights holders/market limitations — looks at whether third parties have a proprietary interest in such rights that may affect the celebrity's ability to engage in endorsement campaigns. For example, Billie Holiday's music rights are controlled separately from her image-and-likeness rights. The ability to market a celebrity like Holiday is thus more contingent on the licensee's securing the necessary clearance to her music. These kinds of transaction costs can impair the value of image-and-likeness rights.

Roesler's fourth factor — past comparable deals — simply looks at past offers or solicitations for the celebrity to engage in an advertising or marketing campaign. A celebrity who indiscriminately pitches himself to companies is generally a celebrity for whose image and likeness there is little demand.

The fifth and final factor (not including the six “rehab” factors) looks at synergy with a company or product. This factor looks at the overlap between the celebrity and the audience to which he appeals. For example, Jim Kelly's image and likeness would have benefits to marketing Charlie the Butcher's beef on weck to Buffalo Bills fans, but might not be nearly as valuable in plugging Dunkin' Donuts to Patriots fans in Boston.

[*111] With all of these factors in mind Roesler went to work to determine Jackson's posthumous marketability. This meant he had to balance Jackson's undoubted gifts as an entertainer against the stigma he bore as an accused child molester. Roesler concluded that the effect of the abuse allegations outweighed reputational rehabilitation during Jackson's life. Sensationalist media coverage heightened and prolonged this damage to his marketability. And this stigma is reflected in his lack of endorsements or merchandise agreements unrelated to a musical tour or album from 1993 until his death.

Looking at Jackson's Q score adds some social-science weight to this obvious conclusion. Before 1993 Jackson was at least as popular as other performers in similar categories. By 1995 his positive Q scores started to decline. And after 2000, Jackson had an overall lack of likeability that continued through 2006. There was not even a recorded Q score for Jackson from 2007 and 2008 — which is itself a bit of contemporaneous proof that Jackson's image was so bad that companies felt it unnecessary or undesirable even to ask that it be studied.

Image-and-likeness rights of a musical celebrity are very closely tied to the ability to license his music. Jackson's compositions and master recordings are protected under federal copyright law. A marketer who wanted to combine Jackson's image and likeness with his music would presumably have to pay more [*112] if songs associated with Jackson, such as “Thriller”, were owned even in part by others.

Roesler viewed Jackson's inability to consummate marketing deals as evidence of his weak marketability after the scandals. In the 1980s Jackson regularly received third-party inquiries for licensing. But after 1993 they dwindled and then virtually disappeared even as Jackson remained a popular musical artist. Though the HIStory World Tour was a success, Sony couldn't recoup a $6 million refundable payment that it made to Jackson for merchandise royalties. And even more recently, the This Is It tour — despite an almost instant sale of 360,000 tickets — did not enable the tour promoter to make a merchandise agreement before Jackson died.

Roesler nevertheless considered the possibility of a foreseeable postdeath boom. This is a well-known phenomenon in the industry; the death of even a faded celebrity can generate extraordinary media attention and a corresponding spike in revenue for his products. Such spikes are routine but temporary, and we agree with the witnesses who said they would fade away after three to five years.

Roesler thought Jackson's potential postdeath boom would last five years. When calculating this amount, Roesler could not ignore Jackson's inability to rehabilitate his image. Jackson's image was so bad that despite many sold-out [*113] shows from the This Is It tour, Jackson generated only $24 worth of image-and-likeness revenue during the last six months of his life. Roesler concluded that Jackson's image-and-likeness rights in the first year after his death should not be more than $2.5 million.

Roesler used this $2.5 million base to estimate the growth and decline of Jackson's postdeath image-and-likeness revenues. Roesler did not believe that consistent growth over five years was reasonable, but did conclude that some growth would occur in years 2 and 3 after death — with the highest revenue occurring in year 3 — and then begin to decrease. Roesler estimated that Jackson's projected image-and-likeness rights revenue in the 10 years following his death would total $22.25 million, an amount far in excess of the image-and-likeness rights revenue Jackson generated in the decade before he died. The following table summarizes Roesler's 10-year revenue projection:

[*114] Year after death	Estimated projection
Year 1	$2,500,000
Year 2	2,750,000
Year 3	3,000,000
Year 4	2,750,000
Year 5	2,500,000
Year 6	2,250,000
Year 7	2,000,000
Year 8	1,750,000
Year 9	1,500,000
Year 10	1,250,000

Expenses. Image and likeness is no exception to the rule that “you gotta spend money to make money.” Once Fishman, with Roesler's help, had a 10-year projection of revenue, he could estimate expenses and derive a final value. These expenses were for commissions and costs such as accounting, legal, and other common business-administration fees. Fishman estimated that due to an unusually high number of poachers of Jackson's image and likeness, there would likely be additional legal expenses.

Fishman estimated these expenses as follows:

• Commissions: 35% of revenue based on Roesler, discussions with others, and his expertise. [*115]

• Public relations: $20,000 monthly ($240,000 annually) to rehabilitate Jackson's image.

• Business-administration expenses: 5% of revenue which would include accounting, management, general/routine legal, and other general overhead.

• Infringement-related legal fees: 20% of revenue which would include legal fees for the protection of the legal rights associated with Jackson's image-and-likeness rights, as well as his ongoing cost to combat infringements of those rights.

This led to total estimated expenses over the next 10 years of $15.75 million. The following chart summarizes both revenue and expenses:

Year	Revenue	Expenses	Projected pretax income
1	$2,500,000	$1,740,000	$760,000
2	2,750,000	1,890,000	860,000
3	3,000,000	2,040,000	960,000
4	2,750,000	1,890,000	860,000
5	2,500,000	1,740,000	760,000
6	2,250,000	1,590,000	660,000
7	2,000,000	1,440,000	560,000
8	1,750,000	1,290,000	460,000
9	1,500,000	1,140,000	360,000
10	1,250,000	990,000	260,000

[*116] He projected the resulting pretax profit for Jackson's image-and-likeness rights at $6.5 million.

Tax Affecting. The next step in Fishman's analysis was to tax affect this income stream. Fishman believed that the buyer would be a C corporation because C corporations have bought other celebrity image-and-likeness rights, and C corporations would have the capital for all the expenses associated with exploiting Jackson's image-and-likeness rights. Fishman used a 35% tax rate in his analysis. The following table shows after-tax income:

Year	Pretax income	Tax rate	Postax income
1	$760,000	35%	$494,000
2	860,000	35%	559,000
3	960,000	35%	624,000
4	860,000	35%	559,000
5	760,000	35%	494,000
6	660,000	35%	429,000
7	560,000	35%	364,000
8	460,000	35%	299,000
9	360,000	35%	234,000
10	260,000	35%	169,000

Present Value. Fishman then discounted these projections to present value using the WACC. First, Fishman determined his cost of debt. After reviewing [*117] market interest rates for public companies, Fishman used 7.5%. Taking into account a 35% tax rate, the cost of debt was 4.88%.

Fishman next needed to determine the cost of equity. Unlike the other experts, Fishman used two methods to calculate the cost of equity: the build-up method and the capital-asset pricing model (which we described supra p. 68). The build-up method calculates the cost of equity as the sum of the risk-free rate of return, the equity-risk premium on small publicly traded common stocks, and the specific company-risk premium.

Fishman first determined the cost of equity under the build-up method. He started with a risk-free rate of 4% based on U.S. Government bond rates for 10, 20, and 30 years. He then added the equity-risk premium on small publicly traded stocks and specific company-risk premium. The studies he used showed that the premium on the largest and least risky public companies ranged from 1.5% to 7% while smaller, riskier firms ranged from 9.7% to 13.1%.³⁹ Fishman reviewed these studies and believed that the equity-risk premium of firms similar in size and risk to Jackson's image and likeness ranged between 10.3% and 15%. Based on his own analysis of these ranges, he chose an equity-risk premium of 12%.

[*118] Due to what Fishman thought was additional risk related to Jackson's image and likeness — the history of erratic revenue, the need to rehabilitate his image, the expectation that there would be a large number of infringers, and the risk associated with the industry — Fishman added a risk premium of 2%. This led to a total cost of equity of 18% under the build-up method.

Under the capital-asset pricing model, Fishman used the same risk-free rate of 4%. Based on his analysis of empirical studies from periods through 2008, Fishman selected a market-risk premium of 6%.

Fishman used three different methods to calculate beta and ended up with a range of 1.3-1.5. This created a total cost of equity under the capital-asset pricing model between 15% and 20%. After considering the results from both methods, Fishman selected a cost of equity of 18%.

Fishman decided on a capital structure of 20% debt and 80% equity after he reviewed selected companies that had debt ranging from 0% to 35% with a median of 20%. With this capital structure, Fishman calculated a discount rate of 15.4%.

We summarize his calculations:⁴⁰

[*119] Year	Projected postax income	Discount rate	Present value
1	$494,000	15.4%	$459,858
2	559,000	15.4%	450,924
3	624,000	15.4%	436,184
4	559,000	15.4%	338,604
5	494,000	15.4%	259,299
6	429,000	15.4%	195,131
7	364,000	15.4%	143,471
8	299,000	15.4%	102,124
9	234,000	15.4%	69,257
10	169,000	15.4%	43,344
Total	N/A	N/A	2,498,196

For years 11-70, Fishman decreased after-tax income by 5% per year and discounted the cashflow to its present value. The remaining 60 years of cashflow discounted back to the date of Jackson's death totaled $217,000. This brought the [*120] present value of future cashflow from Jackson's image-and-likeness rights to $2.7 million.

Fishman's last step was to amortize the tax savings over 15 years.⁴¹ Using a tax rate of 35% and a discount rate of 15.4%, the tax-amortization benefit was $363,000.

The bottom line — an opinion that Jackson's image and likeness was worth a bit more than $3 million:

PV of cashflows	$2,715,000
Tax amortization benefit	363,000
Total value for image and likeness	3,078,000

2. The Commissioner

Anson, the Commissioner's expert, took a wildly different approach that led him to conclude that the value of Jackson's image and likeness was just over [*121] $161.3 million. What caused this difference? Anson disagreed with Fishman on both the discount rate and the projected future revenue stream.

i. Discount Rate

To calculate the discount rate using WACC, Anson began by determining that the cost of debt was 3%. He adopted this from the cost of debt for a corporation named CKX, which had bought Elvis Presley's image and likeness. Anson used a risk-free rate of 4.3%, based on the 20-year U.S. Treasury yield. For his market-risk premium, he used 6%, relying on Duff & Phelps.⁴² And for his beta, he again relied on CKX and used its historical beta of 1.68.

Anson also used CKX's capital structure of 28.7% debt and 71.3% equity. These inputs led him to a discount rate of 11%. Here's a summary of the differences:

[*122] Input	Fishman	Anson
Cost of debt	7.5%	3%
Risk-free rate	4%	4.3%
Market-risk premium	6%	6%
Beta	1.3-1.5	1.68
Cost of equity	18%43	14%
Capital structure	20% debt 80% equity	28.7% debt 71.3% equity
Discount rate	15.4%	11%

ii. Revenue Projections

Anson also projected a much bigger revenue stream. Unlike Roesler, he didn't start with the income Jackson himself had earned from image and likeness before his death. Anson instead used what he called “foreseeable opportunities” that he said were reasonably expected at the time of Jackson's death and that would generate revenue from use of his image and likeness. These “opportunities” were:

• themed attractions and products,

• branded merchandise,

• a Cirque du Soleil show, [*123]

• a film, and

• a Broadway musical.

Anson believed that these projects were potential opportunities that a rational investor could implement and that would create revenue attributable to Jackson's image and likeness. They also bear some considerable resemblance to deals the Estate, under its competent management, did do in the years after Jackson died. Though Anson carefully said that he didn't rely on events after Jackson's death in his valuation, he did look at them to assess the reasonableness of his projections. For each of these opportunities, Anson tried to show how those opportunities were foreseeable at the time of death and then projected revenue streams from each.

Let's look at each.

Themed Attractions and Products. Anson believed a themed attraction was forseeable for several reasons: Jackson had discussed the idea of buying the Spanish Gate Drive estate in Las Vegas and making it into both a residence and museum. Anson also thought that Neverland Ranch might become a tourist attraction. He thought there was a possibility of overseas parks or some other kind of attraction because Jackson had thought about two such projects before he died — including an amusement park and music schools in Bahrain, though neither ever [*124] went anywhere. He also thought there might be video-game revenue because Sega had released a Jackson-themed video game back in 1988.

To project a revenue stream from these possible projects, Anson looked to the experience of Presley's estate, which had cut its own deal to develop an Elvis-themed attraction. That deal provided guaranteed minimum⁴⁴ royalties of $169 million and a buyout clause of $450 million. He assumed that the first year after Jackson's death there would be no revenue from a themed attraction, but in 2011, royalties would be equivalent to the year 1 minimum. He assumed that annual royalties would increase at 3.22% through 2019 and then decline at 5% annually, a lower rate than for Presley's estate (whose royalties were to double after three years and then increase by 22% for the next three years). After he plugged in these numbers, he came up with a present value of revenues from themed attractions and products of a bit less than $50 million. For the remaining 60 years Anson chose a terminal growth rate of zero (which means that the themed attraction would produce the same revenue from years 11 to 70). That discounted to a bit less than $37 million.

Total revenue: nearly $87 million.

[*125] Branded Merchandise. Anson believed that revenue from branded merchandise was foreseeable, because Jackson had earned money from it while he was alive, and because his ex-manager Tohme had worked on deals for branded merchandise in 2008 and 2009. Anson again began by looking away from Jackson's actual experience and instead analyzed six comparable agreements in which a celebrity licensed his image and likeness for use on branded merchandise.⁴⁵ Five of these six celebrities were still alive, but Anson believed their agreements were still comparable because the advantages and disadvantages of a living celebrity offset each other. The celebrities on Anson's list were each entitled to guaranteed annual minimum royalty payments and Anson used that as his baseline.

Anson assumed that a hypothetical buyer of Jackson's image and likeness would begin to make money from branded merchandise in 2010. Anson concluded that revenue would be about $1.3 million in royalties at the start, a figure he derived from the average minimum royalty from the comparable agreements of his comparable celebrities. He then assumed that revenues would increase at the industry growth rate of 3.27%.

[*126] This created total revenues for the first 10 years following Jackson's death that were discounted to a present value of $8.8 million. For the remaining 60 years, Anson — again using a terminal growth rate of 0% — concluded total discounted revenues would be $5.7 million. His total projected revenue stream from branded merchandise was $14.5 million. As often seemed to be the case in Anson's analysis there was some congruence between what he said was foreseeable as of the date of death and what the Estate had been able to accomplish — in 2009, AEG and the Estate reached deals with Bravado International that were expected to produce a total of $15 million in income. See supra p. 45.

Cirque du Soleil. Anson also believed a Cirque du Soleil show was possible at the time of Jackson's death because Jackson had already begun negotiations before he died, at least if one believes Tohme and Wishna. So he projected revenue related to a Cirque du Soleil show. He started with data from similar shows. These similar shows had total gross revenues per year ranging from $35 million (1st quartile) to $48 million (3d quartile) with an average of $46 million. Anson projected that gross revenue from a Jackson Cirque du Soleil show would be equivalent to the third quartile of comparable productions and would run for a total of six years. While this was higher than average, it was 20% less revenue per [*127] year than earned by the Beatles show, and over a run shorter than the Beatles show.

Anson projected that revenue for this hypothetical Cirque show would increase at only 3.22% a year — equal to the long-term inflation rate. As compensation for licensing Jackson's image and likeness, Anson believed that the Estate would receive 3.5% of weekly box office receipts as well as 10% of the net profits. Anson took these numbers from the actual agreement struck between the Estate and Cirque du Soleil after Jackson had died. To determine operating profit, Anson used Cirque du Soleil's overall operating profit margin of 23.9% as of 2008.

Anson asserted that there was other image-and-likeness revenue that would arise from a Cirque du Soleil show. The actual agreement between the Estate and Cirque du Soleil established royalty pools of 5% of gross weekly box-office receipts for music-publishing fees and 5% for “music masters” fees. Anson therefore concluded that the Estate would be entitled to a portion (specifically 74%) of the music-publishing royalty pool. We have no idea why he included this income, derived from Jackson's copyright assets, in his analysis of possible revenue from Jackson's image and likeness.

[*128] All of this income from a projected six-year run of the Jackson Cirque du Soleil show when discounted to present value yielded a bit less than $19 million. Anson pointed out that the actual Cirque du Soleil show produced gross sales 2.5 * his projections.

Film. Anson believed that revenue from a film was foreseeable because Jackson owned archives of footage assembled over his 45-year career in the entertainment industry. Jackson also had the idea before he died of becoming a movie star and had released videos in the past.

Anson therefore projected revenue from a potential Jackson film. He began with a list of 13 films he asserted were comparable and broke down their revenue streams into four kinds of earnings: theater-ticket sales, video sales and rentals, television licenses, and ancillary sales. Anson included both documentary films and biopics on his list. And he made no effort to derive what portion of the revenue from the movies on his list came specifically from use of a subject's image and likeness.

The gross receipts from the films on the list ranged from $245 million (8 Mile which featured Eminem) to $13 million (Why do Fools Fall In Love? which featured Frankie Lymon). He reduced this gross income by certain expenses — about 50% of gross receipts is on average retained by exhibiting theaters as “film [*129] rental.” Anson assumed a 45% film-rental rate for domestic box-office receipts and 55% for foreign box-office receipts. Other expenses include distribution fees, distributor overhead costs, advertising costs, and print costs.

Gross revenue from the video release of Anson's comparables ranged from $262 million (again 8 Mile) to $9 million (again Why do Fools Fall In Love?). There are also a lot of expenses associated with video revenue. These expenses include the retail margin (which Anson projected would be 17%), the rental margin (which Anson projected would be 60% for domestic-rental revenues and 65% for international), distributor fees, advertising expenses, and overhead.

Television and ancillary revenue from the 13 comparable films ranged from $67 million (8 Mile) to $2.5 million (Why Do Fools Fall In Love?). Anson assumed a 30% distribution fee for both television and ancillary revenue.

The 13 comparable films on Anson's list had a wide range of earnings — from a profit of $144 million (8 Mile) to a loss of $41.5 million (The Soloist, which featured Nathaniel Ayers). We summarize these supposedly comparable films:

[*130] Film name	Producer revenue	Production budget	Profit/(Loss)
8 Mile	$184,608,972	$41,000,000	$143,608,972
Get Rich or Die Tryin'	17,189,726	40,000,000	(22,810,274)
Great Balls of Fire!	12,786,248	15,000,000	(2,213,752)
La Bamba	61,282,402	6,500,000	54,782,402
Notorious	34,152,276	19,000,000	15,152,276
Ray	137,468,105	40,000,000	97,468,105
Selena	34,586,494	20,000,000	14,586,494
Shine	57,051,793	5,500,000	51,551,793
The Doors	53,808,846	40,000,000	13,808,846
The Soloist	18,542,205	60,000,000	(41,457,795)
Walk the Line	122,952,511	29,000,000	93,952,511
What's Love Got to Do With It	51,712,231	15,000,000	36,712,231
Why Do Fools Fall in Love?	-0-	7,500,000	(7,500,000)

These are film revenues, not revenues from image and likeness, and Anson did make some adjustments. He projected two sources of income: revenue for licensing content for use in the film and a back-end percentage of the producer's net profit. Anson believed that compensation for Jackson's image and likeness would be 50% of the film's licensing budget, since most footage in any Jackson [*131] film would be derived from footage that the Estate owned. Anson also thought that revenues from the film would be equivalent to the maximum amount generated by the most popular of the 13 comparable films — i.e. to 8 Mile — because of Jackson's fame.

With all these assumptions, Anson projected:

• net theatrical revenue of $2.1 million;

• net video revenue of $134 million;

• net television revenue of $45 million; and

• ancillary revenue of $2.2 million.

Anson then projected a net profit to the producer of $144 million and total compensation to the Estate of $46 million. Discounted to present value, Anson projected it was reasonably foreseeable that the Estate would earn $38.1 million from a film. He carefully noted that the Estate actually received $200 million from Michael Jackson's This is It.

Broadway Musical. Anson also believed that a Broadway musical was foreseeable because in October 2008 Jackson released a written agreement with Nederlander Presentations to develop a musical stage play. See supra p. 29.

Anson computed revenue by looking to other musicals in the jukebox genre — shows based on a musician's compositions, not his life. Anson used all [*132] 12 jukebox musicals that opened on Broadway between 2000 and 2009.⁴⁶ The shows' “trimean”⁴⁷ weeks on Broadway was 106, average percentage of capacity was 73.24%, average ticket price was $98.74, and average performances per week was 7.9.

Anson said he used this data to estimate projected revenues from a hypothetical Jackson jukebox musical. He adjusted the data, however, to reflect the terms of the actual agreement that the Estate later reached with Nederlander. Anson projected the show would run on Broadway for 106 weeks. Under this allegedly “foreseeable” deal — which was based on the actual deal struck after Jackson's death — Jackson would receive 10.37% of weekly operating profits before Nederlander recouped a projected $20 million production cost and 11.86% afterwards. Anson also concluded that Jackson would receive an additional 10% of the net profit earned by Nederlander once the cost of production was recouped. This would result in the Estate's receiving royalties of more than $4 million [*133] between 2011 and 2013. Discounted to present value, this becomes a bit more than $3 million.

Summary. The total of all of these revenue streams discounted to present value is more than $161 million:

Opportunity	Present value
Themed attractions and products	$86,897,899
Branded merchandise	14,460,384
Cirque du Soleil	18,776,037
Film	38,101,767
Broadway musical	3,070,958
FMV of Jackson's image and likeness	161,307,045

C. Analysis

1. Commissioner's Expert

We reject Anson's analysis as fantasy. He:

• valued the wrong asset,

• included unforeseeable events in his valuation, and

• miscalculated the assets' value.

i. Wrong Assets

Anson did not value the asset that he should have — throughout his report and testimony he includes assets other than image and likeness in his valuation. He [*134] did correctly cite California Civil Code section 3344.1(a)(1) for a definition of what a celebrity's heirs inherit as a protected image and likeness. But we find that he ignored the big caveat found in section 3344.1(a)(2) — that California gives no protection for plays, books, magazines, newspapers, musical compositions, audiovisual works, and radio or television programs.

Anson doesn't even mention these exceptions. We understand that Jackson's image-and-likeness rights and his music are related, but find it far more likely than not that Anson included in his valuation of Jackson's image-and-likeness rights the value of rights that California law explicitly excludes from its protection.

These non-image-and-likeness rights — especially copyrights — are separate assets. He should have excluded their values from his estimates of potential income from Jackson's image and likeness.⁴⁸ He instead valued entire projects without making this important distinction. And three of the opportunities that Anson lists — a Cirque du Soleil show, a film, and a Broadway musical — unquestionably contain non-image-and-likeness revenue on even a cursory review. [*135] But the problem extends to all five revenue streams that Anson waded into.

Themed Attractions. Anson relied here primarily on a comparable agreement for Presley. That agreement included copyrights, trademarks, patents, and existing intellectual property licenses — all assets distinct from image and likeness.

Branded Merchandise. Anson used contracts for six comparable celebrities as the bases for his projections of revenue the Estate could expect from branded merchandise. Anson, however, ignored that these contracts required personal services as well. Jackson used to have such contracts, but Anson never explained how Jackson could continue to earn income from them after he died. And three of what Anson called comparable agreements — those for Tony Hawk, Paris Hilton, and Jennifer Lopez — were actually licenses for the use of a celebrity's trademarks, not for the use of image and likeness. One of the agreements even expressly excludes the use of any publicity rights.⁴⁹

Cirque du Soleil. We also find that a Cirque du Soleil show based only on Jackson's music is likely excluded under California Civil Code section 3344.1(a)(2). Even if it were possible that some of the revenue from such a show [*136] was properly attributable to Jackson's image and likeness, it was also attributable to the valuable copyrights that Jackson had in his compositions. Anson completely failed to distinguish the value of these two rights. Anson expressly included additional royalty pools attributable to music-publishing fees — he stated that since the Estate's average ownership of songs is 74%, the Estate would be entitled to receive 74% of the music-publishing royalty pool. This is completely irrelevant to a valuation of Jackson's image and likeness.

Film. Anson's valuation of possible revenues from a film also involved assets other than image and likeness, yet he failed to provide any information on the types of assets licensed in each film. In 8 Mile, for example, Eminem contributed not only his image and likeness, but also his writing, acting, singing, and music — all of which likely affected his compensation. Even for films involving dead musicians, Anson assumed that image-and-likeness rights were required and were actually secured for the movie to be produced. Anson's failure to even identify this issue shows that he was not aware of, or at the least did not take into account, the limited scope of image-and-likeness rights.

The Commissioner would have us value Jackson's image and likeness to include U.S. trademarks, state or common-law trademarks, copyrights, licensing [*137] rights, endorsement rights, franchising rights, and international trademarks. That's not how that right is defined by California law.

We note that Anson's overbroad description of the asset he was valuing was conscious. We have his earlier draft report, in which he based his valuations solely on the image-and-likeness rights defined by California law. We conclude from this that he tried to reach a higher number by broadening the rights he valued.

This is not persuasive.

ii. Unforeseeable Assets

It was also not Anson's only big mistake. We also find that Anson included revenue streams that were unforeseeable at the time of Jackson's death. Even if these potential revenue streams were traceable to Jackson's image and likeness, they were not foreseeable when Jackson died, which means we should not include them in the Estate's gross value.

Section 20.2031-1(b), Estate Tax Regs., defines the fair market value of an estate as limited to what a buyer and seller would have reasonable knowledge of at the time of a decedent's death. Caselaw tells us to hypothesize a buyer who is reasonably informed and who asks a hypothetical willing seller for information not publicly available. Estate of Kollsman, 113 T.C.M. (CCH) at 1175.

[*138] Foreseeability can't be subject to hindsight. Gallagher, 101 T.C.M. (CCH) at 1706. Four of the five revenue streams that Anson included in the valuation of image and likeness were unforeseeable at the time of death.

There is also the problem — entirely foreseeable at the time of Jackson's death — that his reputation as a person (not as a musician) was not in a good place. In the last 10 years of his life, he received almost no revenue related to his image and likeness despite being one of the most well-known persons on Earth. This is all to say that any projection of revenue from the use of Jackson's image and likeness — a person who the last two years of his life was so unpopular that he did not even have a Q score — should be met with skepticism. Anson simply glossed over Jackson's having been accused multiple times of the most heinous acts in his analysis of each supposedly foreseeable revenue stream.

Themed Attractions. The first was themed attractions and products. Anson stated this revenue stream was foreseeable because Branca boasted two months after Jackson's death that his image and likeness rivaled Presley's for earning power. We find these statements a wholly inadequate justification for projecting that anyone could reasonably foresee a Jacksonian Graceland — much less a themed hotel, golf courses, resort, residential developments, and commercial developments such as casinos, restaurants, spas, gyms, lounges, clubs, retail/ [*139] merchandise outlets, and entertainment attractions. The only predeath whisper of a hint that this was a possibility was a few of Jackson's passing thoughts that maybe he would buy a Las Vegas estate to operate as a museum during life and after his death it might become like Graceland.

There is a bit of evidence that Jackson had twice unsuccessfully tried to establish overseas entertainment attractions, but we find that this is not enough to make any kind of theme park foreseeable. Jackson was close to financial ruin at the time of his death. Though the successful ticket sales to the This Is It tour suggest the possibility that he was starting to get out of debt, the idea that he or his Estate could purchase a large property to be run as a museum any time soon is ridiculous.

This leaves Neverland as the sole place possible for a theme park and attraction. Anson also believed that Neverland could be used for this theme park. To any reasonable observer, however, Neverland was more of a recent crime scene than a future wonderland because of the stigma associated with the child-abuse allegations. Common sense suggests that a home owned by an alleged child molester where the alleged molestation took place would be less than an ideal spot for a theme park for children. There was also an issue as to whether, at the time of Jackson's death, Neverland Ranch would even be available to the Estate. Jackson [*140] didn't have any personal possessions at Neverland when he died. And Neverland was owned by an LLC whose members included both Jackson and Colony Capital. Without Colony Capital's consent, the property could not be used for anything.

Branded Merchandise. The foreseeability of branded-merchandise sales is also a problem. Anson relied on a few predeath events to conclude that a merchandising agreement was foreseeable. Jackson was in fact in talks with Bravado about a potential merchandising deal related to the This Is It tour. But at the time of death there was no final, or even imminent, deal. This was not for lack of effort on the part of Jackson and his representatives.

Jackson's reputation at the time of his death has to play a major role in any findings about the foreseeability of a merchandising agreement as an asset, much less one that can be attributed to his image and likeness. Historical data shows that in the years before he died, Jackson had no merchandising deals despite his success in selling his music. Anson fails to adequately acknowledge this fact. Even though a great many people still liked Jackson's music, many fewer liked him as a person — which as we learned from credible testimony is a driving force behind the purchasing of branded merchandise, see supra pp. 32-33 — and which caused him to not be able to market his image and likeness.

[*141] Cirque du Soleil. Even a Cirque du Soleil show was far from foreseeable when Jackson died. Anson relied on an email that described a meeting between Jackson's representative and purported Cirque du Soleil executives that was written before Jackson died. No such meeting ever occurred — Wishna only pretended to be a Cirque du Soleil representative. At no point before Jackson's death, or immediately afterward, were there any discussions of a possible Cirque du Soleil show. We find that projected revenues from a Cirque show are not part of the Estate's value.

Film. Jackson's contract with AEG Live did give AEG the right to produce three feature films. Jackson said two weeks before he died that he was interested in making a feature film. He apparently spoke at one time of wanting to release one movie per year for the next five years. Anson, however, assumed that there was existing footage of the rehearsals for the This Is It tour that could be used to produce a wide-release film.

We find, however, that this does not make a potential film foreseeable. Jackson did leave behind a lot of rehearsal footage. Let us even assume that some part of the revenue from such footage could be attributable to Jackson's image and likeness rather than copyrighted compositions and copyrightable performances. There would still be the problem that the footage was very raw, and needed to be [*142] cooked into a film by other hands; the rights to the footage needed to be established, which was done only by the distinct talents that the managers of Jackson's estate brought to the project on their own.

And there is the same problem that lurks everywhere in our analysis of the value of Jackson's image and likeness — his poor reputation other than as an entertainer. Any projection that finds a torrent of revenue, and not just a trickle, from such a man's image and likeness — especially one who in the last two years of his life was so unpopular he did not even have a Q score — is simply not reasonable.

Anson ignored this rather severe limitation.

iii. Faulty Calculations

The last problem with Anson's report is its math.

Estate Expenses. Anson failed to incorporate any expenses associated with the management of Jackson's image and likeness. CKX — the corporation that Anson relied heavily on throughout his analysis — had expenses that were 69.9% of revenue in 2008 and 68.1% in 2007. Anson incorrectly failed to include such necessary expenses in his future cashflow projections.

Discount Rate. Anson also failed to take a more holistic approach when he was determining the discount rate. In his analysis, Anson looked only to CKX to [*143] determine the discount rate. It is fair to ask whether CKX is a fair representative of the entire market.

There are some notable differences between CKX's business and Jackson's image and likeness. One big difference is that CKX had other sources of revenue such as American Idol, its attendant music rights, and a talent agency. It also makes money from licensing and merchandising, music and television, artist management, themed attractions, and live events. These revenues are not small and are a much bigger part of its income than revenues from celebrities' images and likenesses. Relying on CKX alone for a discount rate was a misstep.

There were also some serious problems with Anson's analysis of the individual opportunities he chose to value.

Themed Attractions. Anson relied too heavily on the Presley agreement for guaranteed minimum royalties in his calculations of projected revenue from themed attractions. He projected $169 million in royalty payments.⁵⁰ But the Presley estate earned only $9 million from that contract — the first guaranteed payment (which was received late). The deal was a flop — by March 2009 it was terminated with no further payments. Using the full potential value of this [*144] agreement — of which only $9 million was ever actually paid — as the basis for projections of foreseeable revenues of Jackson's own image and likeness was unreasonable.

Branded Merchandise. The celebrities Anson used to calculate foreseeable revenue from branded merchandise were anything but comparable. Five of the six were still alive. Anson asserts that the advantages and disadvantages of a live client net out — but that can't be true. Living celebrities can package their images and likenesses with personal service to increase their marketability — an advantage denied to the dead. And Anson completely ignored Jackson's lack of merchandising deals for years before he died.

Cirque du Soleil. Anson relied on five comparable shows for his projections of this revenue. This was misleading. At the time Jackson died, there had been only one Cirque du Soleil show built around the music of an artist--The Beatles LOVE show.⁵¹

Broadway Musical. Anson's revenue projections from this “opportunity” fails to address how overall revenue of a show relates to image-and-likeness rights. Anson relies on the actual Nederlander agreement (which is itself a [*145] problem since its terms were reached only after Jackson died), but that agreement specifically stated that the royalty rate owed to the Estate was for Jackson's composition rights — not use of his image and likeness. Anson didn't even mention this, but included these streams in his projections.

Anson's analysis — quite apart from the taint of his perjury — is unreliable and unpersuasive. It doesn't value the correct asset, it includes unforeseeable opportunities, and it reflects faulty calculations.

2. The Estate's Expert

Fishman's analysis — though we don't agree with it in all respects — is much closer to reality. Fishman gave proper weight to the effect that the allegations had on Jackson's ability to market his image-and-likeness rights during his lifetime up until his death. He recognized that, despite selling tens of thousands of tickets for the This Is It tour, Jackson was unable to successfully market his image and likeness even for tour merchandise.

We do find it reasonable for Fishman and Roesler to look not only at the 10 barren years that preceded Jackson's death. By going back 30 years, but weighting recent years more heavily, Fishman and Roesler gave a nod to the fact that Jackson was one of the most well-known people on earth while also [*146] recognizing that Jackson could not market his image and likeness during the last years of his life.

We are also satisfied with Roesler's confirmation of his revenue projection. Despite Jackson's reputation, Roesler projected its value as second only to that of Marilyn Monroe. We agree with Roesler that at the time of death, Jackson's rights were not worth more than hers. Though Jackson was more famous than Monroe, Monroe's reputation was unburdened by allegations of child abuse. Anson took issue with these comparisons, but Roesler used them simply as a check, not as primary data for his revenue projections.

We also appreciate Roesler's holistic analysis of revenue projections. Roesler considered factors beyond historical data such as Jackson's reputation, his Q score, and the limits of image-and-likeness rights. We find that his analysis accurately captured many of the issues surrounding Jackson's image and likeness. The Estate's experts also accounted for a postdeath spike, which because of the marketability issues had a limited effect. We agree with Roesler that it is reasonable to predict that revenues would peak in year 3, then decline at a rate of 10%-17% from years 4 to 10, and then decrease 5% every year for the remaining life of Jackson's statutory image-and-likeness right.

[*147] We also agree with Fishman's estimates of the future expenses of managing Jackson's image-and-likeness rights. They too are reasonable, and we have no cause to stray from his expert opinion on this point either. (The one major adjustment we make is the discount for tax affecting. As we discussed above, we will not tax affect in this case. See supra p. 82.)

Our only other quibble is with Fishman's method of discounting future income back to present value. For the amounts projected in years 11-70, rather than simply projecting each year back to the date of death, Fishman projected revenue from years 11-70 back to year 10. Then in a separate calculation he further discounted this subtotal back to the date of death. Fishman did not explain why he chose to calculate present value this way.

3. Our Calculations

We begin our own calculation.⁵² We start with Roesler/Fishman's 10-year projected revenue stream. We adopt Fishman's estimate of the expenses of administering Jackson's image and likeness. We also agree that the Estate would need to spend a significant amount of money to rehabilitate Jackson's image.

[*148] We project revenue and expenses separately for the first 10 years, then decrease net income by 5% each year for years 11-70. We will not tax affect. We will use a discount rate of 15.4% — the same as Fishman's.⁵³ Rather than discounting years 11-70 back to year 10 and then again discounting this entire amount back to the date of death, we will discount each year back to the date of death. With these adjustments (which include rounding numbers) we conclude that the value of Jackson's image and likeness at the date of death was just over $4.1 million.⁵⁴ To summarize:

[*149] Year	Revenue	Expenses	Projected income	Discount rate	Projected PV
1	$2,500,000	$1,740,000	$760,000	15.4	$707,474
2	2,750,000	1,890,000	860,000	15.4	693,729
3	3,000,000	2,040,000	960,000	15.4	671,053
4	2,750,000	1,890,000	860,000	15.4	520,929
5	2,500,000	1,740,000	760,000	15.4	398,922
6	2,250,000	1,590,000	660,000	15.4	300,201
7	2,000,000	1,440,000	560,000	15.4	220,724
8	1,750,000	1,290,000	460,000	15.4	157,114
9	1,500,000	1,140,000	360,000	15.4	106,550
10	1,250,000	990,000	260,000	15.4	66,683
Subtotal					3,843,378
Years 11-70			Discounted by 5% from the previous year	15.4	310,533
Total					4,153,912

[*150] VII. New Horizon Trust II and Sony/ATV

The second asset at issue — NHT II — is a Delaware trust in which Jackson held his 50% interest in Sony/ATV.

A. The Asset

Sony/ATV's origin lay in Jackson's decision in the mid-1980s to begin buying music compositions, and especially his key decision to buy the ATV Music Publishing Catalog — the catalog that contained at least 175 Beatles songs — in 1985. Ten years later, Jackson had merged this catalog with Sony's publishing business to form Sony/ATV. The operating agreement gave both members a 50% interest and entitled both to equal numbers of appointments to a board that had the power to approve major decisions.

From 1998 to 2007, as we've already found, Jackson borrowed against his interest in this company. Whenever he had to get Sony's approval to use his interest as security, Sony squeezed him with amendments to the operating agreement. The vise grew especially tight in 2006 when Sony conditioned its approval on Jackson's agreeing to sell 50% of his interest (i.e., 25% of Sony/ATV) at a value capped by a formula. In 2007, Jackson transferred his interest in Sony/ATV to NHT II. The trust, after refinancing by Barclays, held $300 million [*151] of debt, and Jackson agreed to redirect his distributions from Sony/ATV to an interest-reserve account.

B. The Experts' Opinions

1. The Commissioner

The Commissioner again relied on Anson, who first valued Sony/ATV as a whole, then calculated the value of Jackson's interest. Anson used both the income and market approaches and gave equal weight to both methods.

i. Market Approach

Anson's market analysis rested on a key decision that he made at the start — to value Sony/ATV as a music catalog, and not a music-publishing company. Once he made this decision, he believed the market approach was viable because at the time of Jackson's death there was an active market for music catalogs. These catalogs are normally traded at a multiple of net publisher's share (NPS). NPS is gross income less administration fees, royalties due to writers, and direct expenses. Anson used NPS in his market-approach analysis.

Market Multiple. Anson's first step was to figure out the market multiple — what multiple of NPS would a hypothetical buyer pay for Sony/ATV. Anson believed that there was an active market around June 2009 for music catalogs despite the lingering effects of the October 2008 panic. He analyzed seven catalog [*152] sales⁵⁵ to determine an appropriate market multiple. The dates of these transactions ranged from 2006 to 2011, and they yielded NPS multiples between 7.81* and 11.84*. After describing these seven comparable transactions, Anson concluded that he should use the median multiple of 10.7.

NPS Value. Anson next needed to compute Sony/ATV's NPS. He first used Sony/ATV's financials from March 31, 2005, through March 31, 2009. He excluded a $32.3 million payment to Sony/ATV that it received in settlement of a class action suit against Napster in the company's 2009 fiscal year. Anson believed this was a one-time payment that a rational investor would not expect to recur, and he excluded its effect from his analysis.

Since the financials ran only through March 31, 2009, Anson needed to estimate Sony/ATV's NPS between April 1 and June 25. He relied on the company's historical growth rate between 2005 and 2007, because he thought the company's unusually high growth in 2008 was an anomaly that was also unlikely to recur. This average was 3.6%. Anson then determined the partial-year growth [*153] rate of 0.8% based on the number of days between April 1 and June 25, 2009 — 86 days. The following chart from his report summarizes his calculation:⁵⁶

(a) Fiscal Year 2009 NPS	$190,216,000
(b) NPS growth rate	3.6%
(c) NPS partial year growth	.8%
(d) Days between April 1 and June 25	86
(e) Fraction of year	23.6%
(f) Estimated NPS April 1 to June 25 (a * e) * (1 + c)	$45,190,969
NPS from June 25, 2008 to June 25, 2009 (f) + (a * (1-e))	$190,588,952

Anson then multiplied the NPS by the market multiple. This gave Sony/ATV a fair market value of more than $2 billion before deductions. Anson relied on Sony's own numbers as of March 2009 to find that Sony/ATV had debt of more than $755 million. Plug them in, and one gets a net enterprise value:

[*154] Sony/ATV NPS	$190,588,952
Market multiple	10.70*
Total Sony/ATV value	$2,038,473,140
Less debt	($755,300,000)
Sony/ATV net value	$1,283,173,140

Jackson's Share of Sony/ATV. Next, Anson needed to value Jackson's 50% share. This calculation was more difficult than just halving Sony/ATV's net value because Anson had to take into account Sony's right to buy Jackson's interest at a capped price. He did so by calculating separate values for the 50% of Jackson's interest that was affected by Sony's option to purchase and for the 50% of Jackson's interest that was unaffected by Sony's option to purchase.

His calculation of the value of the unaffected interest was easy — 25% of the net value of Sony/ATV. To value the other interest, Anson needed to determine Sony's option price. (A hypothetical rational buyer would presumably not pay full price for the part of Jackson's interest if Sony could immediately buy it away at a lower amount.)

Using Sony's own data Anson calculated the option price at $229.7 million. Anson then combined both of Jackson's interests:

[*155] Sony/ATV net value	$1,283,173,140
Estate's unaffected 50% interest	320,793,285
Estate's affected 50% interest (subject to the option cap)	229,675,000
FMV of Jackson's interest in Sony/ATV	550,468,285

NHT II's Value. Once Anson had calculated the value of Sony/ATV, he was easily able to calculate the value of NHT II by adding in the value of the trust's other assets and liabilities.

Assets:
Cash equivalents	$14,837,303
Accrued interest	84
Estate's 50% interest in Sony/ATV	550,468,285
Total assets	565,305,672
Liabilities:
Barclays capital note	300,000,000
Accrued interest	1,604,896
Barclays capital swap	14,893,017
Total liabilities	316,497,913
FMV of NHT II	248,807,759

[*156] ii. Income Approach

Anson also used the income approach — specifically the DCF method. This required him to determine a discount rate and project Jackson's revenue from Sony/ATV.

Discount Rate. Anson calculated his discount rate using WACC. He chose to rely for many of his inputs on the financials of a company named Vivendi, which was an active buyer of music catalogs around 2009. He decided that the cost of debt would be 4.62%. For the cost of equity he used the capital-asset pricing model. He started with the 20-year U.S. Treasury rate of 4.31% as the risk-free rate. He got a market-risk premium of 6% from Duff & Phelps. Anson then took Vivendi's actual historical beta of 1.26. This yielded a cost of equity equal to 11.84%. Anson then plugged these into Vivendi's capital structure of 29.47% debt and 70.53% equity. Anson didn't tax affect. This led to a discount rate of 9.72%.

Revenue Projections. Anson projected future cashflows from Sony/ATV's own historical financials. His starting point was Sony/ATV's NPS as of March 2009. Anson assumed a growth rate in NPS of 3.6%, which was the historical growth rate for Sony/ATV between 2005 and 2007. (Remember that Anson [*157] excluded Sony/ATV's 2008 growth spurt because he thought it unlikely to recur.) With this growth rate, Anson projected Sony/ATV's NPS for 10 years.

He then estimated Sony/ATV's expenses, but in an unusual way because he believed that the company's historical operating expenses were too high compared to those of its peers. Sony/ATV's average cost percentage was 59.3%, while the cost percentage for the four music-publishing catalogs that Anson used as comparables was 46.8%. Anson thought these costs were irrationally high and a hypothetical rational buyer could reduce them to the music-catalog average of 46.8%. From that Anson could project 10 years of EBITDA for Sony/ATV.

Anson then made some more adjustments before he projected free cashflow. He first subtracted future catalog acquisition expenses. Though Sony/ATV didn't treat this amount as an operating expense (looking on it instead as an investment), Anson saw this as a cash expense that he should subtract. Anson estimated this expense at 8.1% of NPS, based on Sony/ATV's average cost of buying catalogs from 2005 to 2009.⁵⁷

[*158] His second adjustment was to add back “talent cost,” which Sony/ATV classified as an operating expense.⁵⁸ Anson viewed this expense as really just an asset that should be on the balance sheet — not as an operating expense on the income statement — and added it back. This adjustment was 7.03% of NPS, and was based on the average talent cost from 2005 through 2009.

Anson was then able to calculate projected future cashflow projections for 10 years. We summarize:

[*160] Plugging in his discount rate of 9.72% led him to calculate a net present value of $768.5 million in cashflow over 10 years.

Terminal Value. Anson projected Sony/ATV's cashflow for only 10 years, but reasonably believed that it would produce income for many years after. This led him to calculate what's known as a “terminal value.” Terminal-value calculations estimate the value of an indefinite income stream beyond a projection of future cashflows. It is used to project future cashflows that are too difficult to predict with precision.⁵⁹ Using a growth rate of 3.2% (the rate of long-term U.S. inflation), Anson calculated that Sony/ATV's terminal value was $929.8 million when discounted to the present. Here's the math:

PV of free cashflows (10 years)	$768,531,353
PV of terminal value	929,847,186
Gross value of Sony/ATV	1,698,378,538
Less: Debt	(755,300,000)
Net value of Sony/ATV	943,078,538

[*161] Jackson's Share of Sony/ATV. Anson then valued Jackson's 50% share. As with his market approach, Anson's calculation was more complex than just multiplying Sony/ATV's net value by 50% because half of Jackson's interest was affected by Sony's option to purchase.

The value of Jackson's unaffected interest was easy — 25% of Sony/ATV's net value. To value Jackson's affected interest, Anson again used the option price. Here's the math:

Sony/ATV net value	$943,078,538
Estate's unaffected 50% interest	235,769,635
Estate's affected 50% interest (i.e. subject to the option cap)	229,675,000
FMV of Jackson's interest in Sony/ATV	465,444,635

NHT II's Value. After Anson determined the value of Sony/ATV, he was easily able to calculate the value of NHT II, by adding its other assets and subtracting its other liabilities:

[*162] Assets
Cash equivalents	$14,837,303
Accrued interest	84
Estate's 50% interest in Sony/ATV	465,444,635
Total assets	480,282,022
Liabilities
Barclays capital note	300,000,000
Accrued interest	1,604,896
Barclays capital swap	14,893,017
Total liabilities	316,497,913
FMV of NHT II	163,784,109

iii. Discounts

Anson considered whether he should further reduce this number with any other valuation discounts. He did not discount for Jackson's lack of control because he concluded that there was no great difference between Jackson's level of control and Sony's since the Estate could still name half the board and the board still controlled major decisions. The option also gave the Estate power to force a sale of its interest.

Anson also refused to apply any illiquidity discount. In his opinion, music-publishing catalogs were regularly traded at the time of Jackson's death.

[*163] iv. Value of NHT II

As his final calculation Anson averaged his two approaches. His final estimate of NHT II's value was $206.3 million.

2. The Estate

This estimate was precisely $206.3 million more than what the Estate said Jackson's interest was worth.

How is there such a radical disparity in these estimated values? The Estate's expert, Alan Wallis, followed a path broadly similar to Anson's. He, too, valued Sony/ATV as a whole, and then valued Jackson's interest in it. Wallis also used both the market and income approaches — though Wallis relied primarily on the income approach and used the market approach only as a reasonableness check.

Since Wallis relied more heavily on the income approach, we will look at that first.

i. Income Approach

Discount Rate. Wallis calculated the discount rate using a WACC. He too calculated the cost of equity using the capital-asset pricing model. His risk-free market rate was 3.95% — same as a U.S. Government 30-year bond as of the date of Jackson's death. He used a market-risk premium of 6.5% derived from numerous [*164] sources, including Duff & Phelps; Ibbotson & Associates/Morningstar, Inc.; and EY US's own calculations.⁶⁰ To determine the correct beta, Wallis looked for comparable companies, but concluded none were so similar that he could just borrow one. He instead used an average of his comparable companies' betas — giving extra weight to what he thought were more similar companies. This gave him a beta that he concluded fell between 0.9 and 1.1. When put in his capital-asset pricing model, this range resulted in a cost of equity between 10.86% and 11.96%.

He decided on a cost of debt in the range of 3.8% to 5.31% based on five-year A to BB+ rated U.S. bond yields⁶¹ as of the valuation date. Wallis incorporated the effect of a federal tax rate of 35% and the New York State rate of 7.1%. For his hypothetical capital structure — the relative share of debt and [*165] equity — Wallis's research found a considerable range in comparable companies. He did not, however, believe a company would be able to be significantly leveraged in the difficult macroeconomic environment of June 2009. He therefore adopted a capital structure with debt between 20% and 30%.

This all led to a discount rate between 9% and 10.5% with a rounded midpoint of 9.5%, not too far away from Anson's discount rate of 9.72%. See supra p. 156.

Revenue Projections. It was in projected revenues that we come to the most clanging disharmony between the experts. Wallis, instead of relying on Sony/ATV's historical financials to project growth as Anson did, chose to use Sony/ATV's own internal projections as his starting point. These projections ran for only four years into the future, from fiscal year 2010 through 2013.

Though these internal projections were his starting point, Wallis still made some tweaks to get to EBITDA. Like Anson, he added back the talent cost expense. Wallis also believed that two other expenses that Sony/ATV classified as operating expenses in its projections were noncash items. The first was a $7.5 million administrative fee. Wallis believed this expense was really a “payment” on an intercompany loan, and he therefore treated it as a financing item and added [*166] it back into EBITDA. The other expense he added back was catalog amortization, because amortization is a noncash expense.

His last adjustment to EBITDA was to deduct “working capital” — an amount that Sony/ATV had not determined was an operating expense. Changes in working capital reflect the difference between advances paid and the rate at which Sony/ATV actually recouped them. Wallis believed Sony/ATV would continue to pay these advances as part of its normal business and then use recouped advances to invest in new artists at a consistent rate.

Wallis made some more adjustments to his adjusted EBITDA to compute projected free cashflow. He first deducted Sony/ATV's projected $50 million annual allocation for buying new song catalogs. Sony/ATV didn't consider this an operating expense, but Wallis saw this cost as “business-as-usual” and not exceptional. He therefore deducted this amount from EBITDA. He thought the same about Sony/ATV's projected “capex” — the company's line-item cost for acquiring new writers. Wallis believed this was also a regular expense that would reduce cashflow.

The last adjustment he made was to reduce EBITDA for 2010 and 2011 by two “[u]nusual non-recurring items” of $10 million and $17 million respectively. These were in Sony/ATV's projections, but with no explanation. Wallis himself [*167] made a “free cash flows” adjustment to the 2010 numbers by subtracting $34 million with no explanation.

But remember that Sony/ATV looked only four years ahead. Wallis had to figure out a way to project revenue and expenses further into the future. For the next six years, he began with the average growth rate from Sony/ATV's own projections for 2012 and 2013. He didn't use an average growth rate, but growth rates for each revenue stream (mechanical, performance, synch, and other) from 2012 to 2013. He then reduced these rates to a long-term growth rate of 2.1% — the long-term U.S. inflation rate from Global Insight⁶² — after 2013.

When he reduced Sony/ATV's growth rate, he was lowering its value, but Wallis also made some adjustments to the data that did not favor the Estate: He reduced acquisition cost over the last 6 years from an annual $50 million to only $25 million. He increased his projection of net revenue in years 5-10 by assuming that Sony/ATV would keep its working capital at 2013's $16.5 million and not increase it. And he also held constant the annual “capex” expense at $1.1 million.

Wallis did tax affect. He reduced Sony/ATV's projected income by an estimated income tax at a federal rate of 35% and a New York State rate of 7.1%. [*168] After these adjustments, Wallis projected future cashflows for 10 years, and finally predicted that a buyer would benefit from a tax-amortization benefit.

Wallis assumed a buyer would treat this purchase as an asset sale and amortize the capitalized assets. This created a range of values for Sony/ATV between $800 million and $1.05 billion.

We summarize Wallis's 10-year revenue projections (in millions):

[*170] Terminal Value. Wallis then calculated Sony/ATV's terminal value using the Gordon Growth Formula⁶³ at fiscal year 2019 with a long-term growth rate of between 1.6% and 2.6%.

Present Value. With free-cashflow projections for 10 years established and a terminal value formula ready to go, Wallis calculated a range of values using his range of long-term growth rates and discount rates. Wallis concluded that the value of Sony/ATV under the income approach was between $745 million and $1,060 million:

[*171] $m	Discount rate
Long-term growth rate		9.0%	9.5%	10.0%	10.5%
	1.6%	$956	$876	$806	$745
	2.1%	1,004	916	840	774
	2.6%	1,060	963	879	806

ii. Market Approach

Like Anson, Wallis also used the market approach in his valuation. Sony/ATV was, at the time of Jackson's death, the fourth largest music publisher in the world. This prompted Wallis to view Sony/ATV as a music publisher and not just a music catalog.

The distinction is an important one because music publishers are operating businesses, not just investment portfolios of songs. They are normally valued like other businesses as a multiple of EBITDA, not of NPS. Wallis agreed with Anson that NPS is an appropriate metric when the asset in question is a catalog (or even a small publishing business). But because Sony/ATV was very large, Wallis believed EBITDA was a better measure for a market approach; and that meant he first needed to determine Sony/ATV's EBITDA as of the date of Jackson's death [*172] and then determine an appropriate multiple of EBITDA that a hypothetical buyer would pay for it.

EBITDA Value. Wallis calculated the Sony/ATV's EBITDA as of the end of its fiscal year in both 2009 and 2010. He used Sony/ATV's actual results for the year that ended on March 31, 2009, and then he projected results to March 31, 2010. From this starting point, Wallis made some tweaks — notably excluding the one-off Napster settlement that Anson had also excluded. This created a historical EBITDA of $89.41 million for fiscal year 2009 and a projected EBITDA of $96.41 million for 2010. Here are the calculations (in millions):

Year	FY 2009	FY 2010
Revenue	$462.7	$466.0
Less royalty exp.	272.5	272.8
NPS	190.2	193.2
Total operating cost	136.0	132.5
Income from operations	54.2	60.8
Plus dep. and amor.	35.1	35.6
EBITDA	89.4	96.4

EBITDA Multiple. Wallis calculated the EBITDA multiple two ways: as a trading multiple and as a transaction multiple. A trading multiple looks at the market capitalization of a publicly listed company (the number of its shares [*173] multiplied by its stock price) divided by that company's own EBITDA. A transaction multiple looks at data from acquisitions of comparable companies and divides the sale price of a target by its EBITDA.

To calculate his trading multiple, Wallis used only one company, Warner Musical Group (WMG). In his opinion this listed company was easily the most similar to Sony/ATV that had publicly traded stock. On the day Jackson died, WMG had a market capitalization of $789.3 million. Figuring out its EBITDA was more complicated.

To determine a control premium (an additional amount a hypothetical buyer would pay to control a company), Wallis considered two transactions of music businesses — the acquisition of EMI in 2007 and the acquisition of Chrysalis in 2010. Using these two transactions, Wallis determined that he should use a control premium of 30%. The net debt at the valuation date was $1,564 million. This gave WMG an enterprise value of $2,590.1 million. The following table summarizes his calculation:

[*174] Market capitalization	$789,300,000
Control premium	30%
Adjusted market capitalization	$1,026,100,000
Plus: Net debt	$1,564,000,000
Enterprise value	$2,590,100,000

Wallis then adjusted this EBITDA a little bit more because WMG, unlike Sony/ATV, was not only a music publisher; it also had a large recorded-music business. Wallis needed to determine what the trading multiple was for just the music-publishing division of WMG, not the entire company.

To do this he looked at WMG's 2008 and 2009 financials and calculated EBITDA for both divisions. He then ratably allocated WMG's corporate overhead to the two divisions on the basis of each division's reported and forecasted EBITDA.

Next, using what he considered an appropriate multiple for the recorded-music division of 4* or 4.5*, he calculated its value. After he did this he could back into the trading multiple of the music-publishing division. The trading multiples ranged from 7.8* to 11.1*. That meant Sony/ATV had a value between $800 million and $1.1 billion.

[*175] Calculation of a transaction multiple was difficult because Wallis couldn't find any similar transactions around the date of Jackson's death. He relied instead on two comparable transactions that took place in 2011 — between 12 and 18 months after Jackson's death. These two transactions had market multiples of 9.6* and 9.9*. This led him to conclude that Sony/ATV's enterprise value was between $858.2 million and $885.1 million.

iii. Sony/ATV Value

After analyzing the value of Sony/ATV under the income approach and market approach, Wallis had a range of values between $745 million and $1.1 billion:

Method	Amount (in millions)
Income approach	$745-1,060
Market approach (trading multiple)	800-1,100
Market approach (transaction multiple)	858.2-885.1

He went to the high end of this range and concluded that Sony/ATV was worth $1.1 billion. He said he went to the high end to take into account any potential premium for the Beatles songs that Sony/ATV controlled. It is, of course, also true that this makes his opinion rhetorically stronger by being a bit unfavorable to the Estate.

[*176] He then started subtracting from this value. First to be knocked off was debt owed by Sony/ATV — a net amount of $579.3 million based on Sony/ATV's June 30, 2009, balance sheet. He also deducted $12.8 million — which is 5% of the fair market value of “Class B interest.”⁶⁴ After these subtractions, Sony/ATV's net value was $508 million.

iv. Jackson's Interest

Wallis could then calculate the value of Jackson's 50% interest. As had Anson, Wallis looked to see how the option cap affected 50% of Jackson's interest. Wallis relied on calculations prepared by Sony that suggested that the cap would affect the value of Jackson's interest only if half of Jackson's equity value was worth $242.1 million. Because Wallis valued Sony/ATV at less than $1.5 billion,⁶⁵ however, the cap has no effect and he made no adjustment related to Sony's option to purchase half of Jackson's interest.

[*177] Wallis then made two discounts: one for lack of marketability and one for lack of control.

Wallis started with $254 million, which was simply half of Sony/ATV's total value. But remember that Jackson had to cede ever larger chunks of his rights under the Sony/ATV operating agreement to gain Sony's approval for his borrowing against its value. Wallis thought these concessions shrank that value by 20%. We can summarize the most important of these factors:

[*178] Factor	Comment
Dividend prospects	Guaranteed annual distributions of $11 million ended September 2011 (and were directed to meet NHT II's interest obligation); thereafter distributions were at Sony's discretion.
Restrictions on transfer	No transfer could occur without Sony's consent unless through the buy-sell provisions, which could be more timely and costly.
Prospects of IPO/Sale of company	No indication at the valuation date that Sony was considering leaving the music business.
Number of identifiable buyers	Limited at valuation date.
Size of stake	While 50% stake included 50% voting rights on the board, influence is limited to approval of certain “major decisions.”
Access/reliability of information	Jackson's interest carries limited rights to management information other than the annual budget.
General market conditions	The global economy was in recession and availability of financing was limited.
NHT II's debt	A purchaser of Jackson's interest would likely have to assume NHT II's debt (this alone caused a 10% discount).

Wallis also discounted Jackson's interest for lack of control. Wallis viewed the limitations put on Jackson's interest as having a large effect on the value of his [*179] interest. To quantify this effect, Wallis used a discount of 15% — a number consistent with Shannon Pratt's Business Valuations: Discounts and Premiums 24-25 (2d ed. 2009).⁶⁶

With a discount of 20% for lack of marketability and a 15% discount for lack of control, Wallis computed the value of Jackson's interest in Sony/ATV:

Jackson's interest pre discount	$254 million
Discount for lack of marketability	(50.8 million)
Discount for lack of control	(38.1 million)
Jackson's interest postdiscount	165.1 million

[*180] v. NHT II's Value

Wallis's last step was to calculate the value of NHT II. Wallis used NHT II's March 31, 2009, accounts to determine the value of its assets and liabilities. He determined there were assets of $14.7 million and debts of $319.1 million.

The math was simple:

Jackson interest in Sony/ATV	$165.1 million
Plus: NHT II's net assets	14.7 million
Less: NHT II's net liabilities	(319.1 million)
Net value of NHT II	(139.2 million)

C. Our Analysis

This is quite a spread, with Anson's bottom line at more than $206 million and Wallis's at negative $139 million (which would make NHT II valueless as of the date of Jackson's death). To find a reasonable value of our own we will, like the parties, look at both the market and income approaches.

1. Market Approach

We begin by finding that Anson looked at the wrong data, because he treated Sony/ATV as a music catalog and not an operating business in the music-publishing industry. Sony/ATV was not just a music catalog. It had employees and corporate infrastructure to acquire and exploit both old and new [*181] songs; it didn't have to pay anyone else to administer the songs it owned, but it did have to constantly invest in acquisition and development. It was successful in doing so — under Sony's management the company had grown into the fourth largest music publisher in the world. And there was one overriding fact: neither Jackson nor any hypothetical buyer could possibly force Sony to change its goal from being a growing publishing company into a slowly liquidating music-catalog business. We do not find Anson credible in his use of NPS as his starting point in valuing Sony/ATV.

Wallis was much more persuasive that EBITDA was a better base on which to build an evaluation of an operating company. But this leads to a different problem — as of June 2009, there was no market for music-publishing businesses. We know this from Wallis's own report, where he credibly states that “no relevant transactions took place around the [v]aluation [d]ate.” With no relevant transactions, we do not think the market approach — even a market approach based on Sony/ATV's EBITDA — is useful here.

We certainly recognize that identifying comparable transactions and tweaking them to fit the facts of the asset being valued is preferred, and we think that Wallis did the best he could do under the circumstances. It's just that the two transactions he found took place over a year after Jackson's death; and are too few [*182] in number and too different in size for any market approach to be reasonable here. (And one of those transactions was a distressed-asset sale, which makes using it even more of a problem.) Wallis ended up with only one company — WMG — to use in his calculation of a trading multiple. WMG, as Wallis admitted, was extremely different from Sony/ATV because it had a large music-recording business.

We therefore likewise find that Wallis's analysis is also unreliable.

2. Income Approach

This leaves us with the income approach. This requires us to calculate a discount rate and future revenue projections. We'll take them in turn.

i. Discount Rate

The experts nearly converged on an appropriate discount rate. Wallis used one between 9% and 10.5%. Anson used one of 9.72%. These competing experts got there in drastically different ways, however, and we'll look at each of the variables in the models they used.

Anson relied on the average interest rate from Vivendi's first-quarter 2009 financials and came up with a cost of debt equal to 4.62%. Wallis started with the average rate on five-year A to BBB+ rated U.S. corporate bonds and came up with a cost of debt somewhere between 3.8% and 5.31%. We find that a reasonable [*183] cost of debt is 4.62%. While we would prefer not to rely on just one company, Wallis's method is less reasonable and gives us too wide a range — we need a number for the equation. And 4.62% is within Wallis's range. As we've already explained, see supra p. 82, we will not tax affect this number.

Cost of Equity. Both parties used the capital-asset pricing model to compute a cost of equity. We adopt Anson's use of the 20-year U.S. Government bond rate for a risk-free rate of 4.31%, rather than Wallis's 30-year rate. There's no explanation of why one is to be preferred to the other, and in the past we ourselves have seemed not to have a preference. See Estate of Jung v. Commissioner, 101 T.C. 412, 441 (1993). But every expert in this case except Wallis used the 20-year rate in some fashion. We find in that consensus some comfort that it is reasonable for us to do so.

To determine the beta, Anson relied on Vivendi's historical beta of 1.26 while Wallis adopted a range between 0.9 and 1.1 based on the median from comparable companies. We will use a beta of 1.26 for our calculations. We would rather not rely on only one company. But as Wallis admitted there were not many comparable companies, so he broadened the search — presumably thereby including noncomparable companies. He also used a five-year average beta for [*184] each of these companies. His own report, however, stated that historical data didn't apply because the Great Recession had so changed market conditions.

For his market-risk premium, Anson relied on Duff & Phelps and came up with 6%. Wallis relied on an internal report which considered Duff & Phelps as well as other publications and other clients of Ernst & Young, which led him to come up with a rate of 6.5%. On this one, we agree with Wallis because his number comes from more than one reliable source.

Capital Structure. To determine a capital structure, Anson again relied on Vivendi's capital structure of 29.47% debt and 70.53% equity. Wallis's hypothetical capital structure again came with a range — somewhere between 20% and 30% is debt. We will use Anson's capital structure of 29.47% debt and 70.53% equity, and note this puts us within Wallis's range.

Tax Affecting. We have already concluded, see supra p. 82, that we will not tax affect.

Final Discount Rate. We summarize our calculations and compare them with the experts':

	[*185] Anson	Wallis	Tax Court
Cost of debt	4.62%	3.8% - 5.31%	4.62%
Risk-free rate	4.31%	3.95%	4.31%
Market premium	6%	6.5%	6.5%
Beta	1.26	.9 - 1.1	1.26
Cost of equity	11.84%	10.86% - 11.96%	12.5%
Capital structure	29.4% debt	20% - 30% debt	29.47% debt
Tax affecting	No	Yes	No
Discount rate	9.72%	9% - 10.5%	10.18%

ii. Revenue Projections

With the discount rate established, we shift to cashflow projections. We first have to decide whether we find it more reasonable to predict the future with the past or with the prediction of a third party — Sony/ATV — that had an interest in getting it right. We think it more reasonable to use Sony/ATV's own projections. Our major reason is that the music-publishing industry was (and has remained) in a state of considerable uncertainty created by a long series of seismic technological changes. Digital downloads, the advent of easy-to-use pirating software, the development of countermeasures to defeat that software, and similar revolutionary changes in the identification and marketing of new singers and songs make historical data about cashflow exceptionally unreliable in this [*186] industry. We think that projections of future cashflow, if made by businessmen with an incentive to get it right, are more likely to reflect reasonable estimates of the short- to medium-term effects of these wild changes in the industry that even experts, much less judges, are unlikely to intuit correctly. We therefore find that Sony/ATV's projections are more reliable and will, like Wallis, use these midrange projections for the first four years as the most reasonable.

Years 1 Through 4 Projections. We begin our cashflow projections just as Wallis did — with Sony/ATV's own internal projections. We reject Anson's call that a rational investor would lower operating expenses to be more in line with other companies. We are not persuaded that any such adjustment would be based on anything other than speculation. We're also unsure how to compare Sony/ATV's operating expenses to other companies' especially when we consider how both parties struggled to find a company comparable to Sony/ATV in their market-approach analyses.

But we do agree with both parties that some adjustments are needed. We agree with both experts that we should add talent cost back to EBITDA.⁶⁷ We agree with Wallis that the administrative fees should be added back because we [*187] think it was more likely a book-only entry of an amount owed by Sony/ATV to Sony for guaranteeing Sony/ATV's debt. We agree that this makes it better classified as a financing cost. We also agree with Wallis's reasoning for why we should add catalog amortization to EBITDA.

We agree that other adjustments are needed to determine free cashflow. We agree that song-acquisition cost would be a business-as-usual expense and should be subtracted. Sony/ATV was created for the express purpose of actively expanding its music catalog and achieving economies of scale in its exploitation. We find credible Sony/ATV's own projections that it would need to spend large amounts to achieve this goal. We will also deduct the 'Capex' line item as well as the working capital amount from EBITDA. Anson doesn't dispute this, and neither shall we.

Wallis included two nonrecurring items that totaled $27 million. We recognize that Sony/ATV included them, but there's nothing like this in the company's audited financials from 2005 to 2011 and no explanation for what they might be. We will not include these expenses in our projections. Wallis made his own adjustments to free cashflow in the first year of projections, but with no explanation. We see no reason to include it either.

That leads to a four-year projection of cashflow and EBITDA:

Years 5 Through 10 Projections. We now need to project cashflows for years 5 through 10. Wallis used Sony/ATV's expected growth rate from year 3 to 4, then gradually reduced it over the next six years to the long-term growth rate equal to the long-term inflation rate of 2.1%. Anson assumes a consistent 3.6% growth based on Sony/ATV's average growth from 2005 through 2007.

[*189] We find Wallis more credible here. We are persuaded that his method of tapering down the growth rate to the broader economy's expected long-term growth rate is more reasonable than constant growth based on pre-recession rates.

Wallis projected costs would grow at an annual rate equal to expected inflation — 2.1%. Anson would increase them at an annual rate of 3.6%. For the same reason we found Anson's growth rate for revenue less reliable than Wallis's, we find it less reliable again. We will use a growth rate of 2.1% for expenses.⁶⁸

Neither party applies this 2.1% growth rate to all costs. Wallis projected the administrative fee to remain consistent at $7.5 million. He also lowered the song-acquisition cost to $25 million annually based on his longtime observations of the industry. Anson disputed only the earlier $50 million acquisition expense (which we already addressed) but disputed neither of these adjustments. Neither will we. Wallis's projection of capex and working capital remain constant during years 5-10. Seeing no dispute from Anson, we also believe these amounts should remain constant.

With this information, we are able to project free cashflows for 10 years, discounted to present value. This following table summarizes our calculations:

[*191] Present Value of 10-Year Projections. Based on our calculations the present value of the future cashflows through 10 years is $458.6 million.⁶⁹

Terminal Value. We agree with both parties that estimates of revenues beyond 10 years are in this case very uncertain and we therefore need to calculate a terminal value. This means we must decide on an appropriate long-term growth rate. Wallis uses a long-term growth rate between 1.6% and 2.6% based on the average growth rates from 2020 to 2039 he adopted from Global Insight. Anson uses a long-term growth rate of 3.2% based on the U.S. rate of inflation, though he does not cite a source.

We find it reasonable to use a growth rate of 2.1% — the midpoint of Wallis's range. Wallis at least gave us a source, and Anson did not dispute Wallis's rate.

With our discount and long-term growth rates in hand, we can calculate the terminal value. Anson does not say which method he uses and failed to show his work in his calculation. So, like Wallis, we will use the Gordon Growth Formula, [*192] which our precedent also supports.⁷⁰ See, e.g., Kardash, 109 T.C.M. (CCH) at 1237.

Our last year of free cashflow is $113 million. Our growth formula is the long-term U.S. inflation rate of 2.1%. Our discount rate is still 10.18%. This makes our terminal value $1.428 billion. The present value of this terminal value is $569 million.

iii. Sony/ATV Value

With our revenue projection and terminal value complete, we calculate that the value of Sony/ATV under the income approach using the DCF method is $1.027 billion.

We next subtract the net debt from Sony/ATV. There was a dispute about this. Anson determined that Sony/ATV's debt was $755 million based on Sony's calculations from March 31, 2009. Wallis determined that it was $579 million based on Sony/ATV's June 30, 2009 balance sheet. We believe the debt should be $579 million. First, the net debt that Wallis used is lower than what Anson used, so we can treat it as a concession. Second, Wallis used the most recent financials available as of the date of death. Sony/ATV's net value:

[*193] PV of 10-year projections	$458,552,174
PV of terminal value	568,515,193
Value of Sony/ATV	1,027,067,367
Less: Sony/ATV debt	579,200,000
Net value of Sony/ATV	447,867,367

iv. Jackson's Interest

If there were no discounts or adjustments, Jackson's 50% of Sony/ATV would be worth $224 million. However, the parties disagree about three potential adjustments:

• the effect of the option price,

• a discount for lack of marketability, and

• a discount for lack of control.

Option Price. Sony had the option to buy 50% of Jackson's interest. The parties calculated two different numbers for the strike price of the option: Anson thought it was $229.7 million and Wallis thought it was $242.1 million. This is another fight we don't need to referee: We've found that the unadjusted value of Jackson's interest was under $225 million. Fifty percent of that interest is well below either Anson's or Wallis's calculated strike price, so we find that the option price has no effect on the value of Jackson's interest.

[*194] Discount for Lack of Marketability. We are also unpersuaded that Jackson's interest lacks marketability. Though Wallis goes to great lengths to point out possible issues with Jackson's interest, we believe they are better reflected in the discount for lack of control. The idea that ownership interest in the company that holds rights to 175 Beatles songs isn't marketable seems like a stretch.

Discount for Lack of Control. We last need to determine whether there should be a discount for lack of control. Wallis believed that there should be a 15% discount because Jackson had only some control of the board and not management, while Anson believed there should be no discount. We believe somewhere in the middle is most reasonable. Jackson's interest is less valuable than Sony's since Jackson couldn't control day-to-day activities. But at the same time, Jackson was still able to block major decisions, which presumably were more important, and he was still a 50% owner. We find that only a 5% discount for lack of control is appropriate, though it might not matter that much.

We summarize:

[*195] Sony/ATV net value	$447,867,367
Jackson's unencumbered 25% share	$111,966,842
Jackson's encumbered 25% share (no effect of option price)	$111,966,842
FMV of Jackson's share pre-discount	$223,933,683
Discount (lack of control)	5%
FMV of Jackson's share	$212,736,999

v. NHT II's Value

Our last step is to value NHT II. The experts disagree slightly about what the net value of NHT II's assets not subject to the Sony/ATV interest are ($301.7 million versus $304.3 million). This difference appears because Anson calculated NHT II's value based on the Estate's tax return, while Wallis based his calculations on NHT II's March 31, 2009 accounts. Though it will have only a minimal affect on the value of NHT II, we will adopt the numbers on the Estate's tax return used by Anson.

Here's the calculation:

[*196] Assets:
Cash equivalents	$14,837,303
Accrued interest	84
Estate's interest in Sony/ATV	212,736,999
Total assets	227,574,386
Liabilities:
Barclay's capital note	300,000,000
Accrued interest	1,604,896
Barclay's capital swap	14,893,017
Total liabilities	316,497,913
FMV of NHT II	(88,923,527)

With a negative value, we find that the value of NHT II on the day Jackson died was $0.

VIII. New Horizon Trust III

A. The Asset

The last asset that we have to value is NHT III, which is also a bankruptcy-remote trust. Much of its value is not in dispute — it has cash of $3.5 million and liabilities of $72 million. But the value of its major asset — Mijac Music — is very much in dispute. Mijac is a music catalog that owns copyrights in [*197] compositions from a variety of artists, most notably Jackson himself.⁷¹ It is administered by Warner Bros. Music. And, in addition to Mijac, Jackson assigned to NHT III the income he earned from his writer's share of performance revenue from BMI.

Mijac is the most difficult of Jackson's assets to value because its income derives from five different groups of songs, and each group produces income from three different sources. And sometimes distinctions must be made between domestic and international income.

The five groups are:

• songs composed and performed by Jackson and released before he died (released Jackson songs);

• songs composed by Jackson but performed by someone else;

• major works by other artists;⁷² [*198]

• minor works by other artists; and

• songs composed and performed by Jackson that were unreleased before he died.

Before we look at how the experts went about their work, we need to review how Mijac earned its income. It is a catalog of music compositions. Composition income comes mostly from three sources:

• mechanical revenue,

• performance revenue paid by BMI (including society rebates),⁷³ and

• synch fees.

The Commissioner again used Anson to value Mijac. The Estate used Dahl, a former principal at Moss Adams and the founder and the current president of Dahl Consulting Group. Both experts concluded that a DCF analysis was the proper method of valuation. We agree.

[*199] This means that the income stream from each of these sources of revenue from each group of compositions needs to be projected and discounted to present value to arrive at Mijac's total value.

B. Mijac's Value and Writer's Performance Royalties

In our valuations of Jackson's image and likeness and of NHT II, we went through each expert's analysis and then completed our own valuation by analyzing where the experts disagreed. Mijac was the most complex and difficult of the assets to value because of the large number of variables in the various sources of revenue and groups of assets.

The easiest variable is the overall discount rate — Dahl's rate was 9.7%, and Anson's only a tiny bit different at 9.72%. We will round Anson's discount rate to 9.7% and use that number.⁷⁴

We're left with the far more complicated problem of deciding on a reasonable cashflow for each income source from each composition group. This is made even more complicated since Jackson's death itself affected some of them, but not all.

[*200] We organize this part of the opinion by focusing on a few points of disagreement that affect several of these groups and sources. The effects of our analysis then reverberate throughout the rest of this section as we look at each of the groups and sources that flow together to create Mijac's revenue stream.

There are three major disagreements between the experts:

• the number of unreleased songs at Jackson's death;

• the correct starting point for the cashflow projections; and

• the spike and growth rate for each group of songs.

We address each separately.

1. Unreleased Songs

The first major disagreement between the experts is how many unreleased songs Jackson had at the time of his death. The experts came up with two very different numbers.

We'll start with Anson. Anson believed that Jackson had an enormous number of unreleased songs on the day he died. At one point in his review he thought there were more than 153, though he noted that some sources hinted that the actual number of unreleased songs could be much higher — possibly over 200. At another point in his report he said that there were 133 unreleased songs at the time of death. But, in any event, he was sure there were at least 105.

[*201] This was at least a bit confusing and may be a consequence of Anson's source material. His primary source was Wikipedia⁷⁵ though he also relied on a deposition Jackson gave back in 1993, a book, Michael Jackson: For the Record, and a 2015 report from the Estate showing previously unreleased songs at the time of death that have since been released. See Campbell ex rel. Campbell v. Sec'y of Health & Human Servs., 69 Fed. Cl. 775, 781 (2006).

He did have some reliable information — the songs that the Estate actually released. Anson concluded there were only 15 such songs but then acknowledged at trial that the Estate had released 21 new songs after Jackson's death.

Anson then claimed that there were another 45 unreleased songs whose existence the Estate either confirmed or did not dispute. He cross-referenced these titles with Michael Jackson: For the Record, which identified 38 of them. Anson's number then grew some more when he included 22 songs from Michael Jackson: For the Record that the Estate claimed there was no record of.

Anson also had a couple of positions on the important question of how much composition credit Jackson would have for these unreleased songs: It was common for Jackson to share composition credit on his songs, and Anson did [*202] some kind of analysis that led him to conclude that Jackson owned 84.2% of these unreleased compositions.

Dahl was nearly as confusing, but at smaller numbers. Dahl believed there were 68 songs that could be validated as compositions written by Jackson and another 11 whose ownership or even existence was uncertain, as well as a larger list of songs with no record of Jackson's having composed them at all. Dahl included in his list 19 songs that the Estate has since released and 62 songs that it has not — which does not add up to 68. Dahl also appears to have accidentally excluded 2 songs that were released posthumously but that he did not list.⁷⁶

The Code tells us that the gross estate includes the value of “all property to the extent of the interest therein of the decedent at the time of his death.” Sec. 2033. The Code does not distinguish between known and unknown property, see id., and we must reject any notion of what amounts to a discount or bonus to Mijac's valuation because of the Estate's initial uncertainty about what assets Jackson held at the time he died. We must be cautious about using information [*203] from later events in valuing his assets; we must be a factfinder about what assets were actually in his estate at the time he died.

We need to find out how many unreleased songs Jackson had at the time of his death, and in which of these songs he qualified for a composition copyright, either as a lyricist or composer. This is a question of fact, not a matter of valuation-expert opinion. It is also not an easy question given the way Jackson worked and the absence of any box in his personal vault labeled “complete unreleased songs, use if needed.” The fact we begin with is that Jackson's vault held 7,000 to 10,000 pieces of tape. The Estate called an entirely credible witness, John Doelp, whose job was to go through these snippets and figure out what was there. Doelp was a longtime employee with Sony and has worked in several capacities including marketing, finance, and artist-and-repertoire (finding talent). While he was running the marketing-and-sales department, he had himself worked with Jackson. Yet Doelp found only 2 completed songs in nearly finished form that could be released, and another 25-30 with full vocals (some of which couldn't be released).

Since Jackson's death, the Estate has released 21 songs and confirmed the existence of 62 other songs that have not been released and that won't be because [*204] they are not of commercial quality.⁷⁷ We also find that, apart from the two songs that were in nearly finished form, the remainder needed considerable work to bring them up to commercial quality. We suspect that if the Estate had pressed the point, their unfinished character would have meant an allocation of some of their value to the Estate's own efforts instead of their value as of the date of Jackson's death.

Though the songs were scattered among 7,000-10,000 pieces of tape, the songs were there. There were at least 83 of them, because the Estate released 21 after Jackson's death and confirmed another 62. We believe Anson's much higher total number of songs is unreliable for several reasons. The sources of his information are unreliable — Wikipedia, an interview that is over two decades old, and a book whose own sources are unclear. Dahl's source, which is the Estate itself, is more reliable here, and on any close question we can rely more on Dahl simply because he didn't create the same problems with credibility that Anson did. This means we find that there was a total of 83 unreleased songs available at Jackson's death.

[*205] Our inquiry doesn't end there, because a song available to Mijac might not be a song valuable to it. For Mijac, we care about composer rights, not performer rights, and we need to determine what percentage of these songs that Jackson had taped were songs that Jackson had composed. Again, since this is a question of fact, we are not limited to information as of the date of death. But the parties disagree here as well — Anson assumed Jackson would get 84.2% of the composer revenue from all unreleased songs, while Dahl gives several percentages that range from 36% to 58%.

We think it is best to begin with the 21 posthumous released songs and determine what composer credit Jackson received on them. The parties stipulated that “This Is It” was cowritten by Jackson and Paul Anka — we find that he got 50% writer's credit. Michael contained 10 songs, 5 of which Jackson wrote or cowrote. Of the 5 songs he wrote or cowrote he received an average writer's credit of 62%. This means that Jackson had composer rights to 31% of the songs on Michael. Xscape contained 8 songs, 5 of which Jackson wrote or cowrote. Of the 5 songs Jackson wrote or cowrote he received an average writer's credit of 57.6%. This gives him composer rights to 36% of the songs on Xscape.

This yields an average composer's credit for Jackson of 34% per song. See infra Appendix A. We believe it is most reasonable from the scant record we have [*206] to use this percentage of composer credit for any unreleased songs. And as we explain later, see infra pp. 213-14, we caution that it would be reasonable for a hypothetical buyer to conclude that most of those unreleased songs had little or no value at all.

2. Starting Point

The next disagreement is the starting point for the DCF projections.

i. The Experts

Anson. For his projection of mechanical royalties for all groups except unreleased Jackson songs, Anson began with average annual royalties from 2007 to 2009. For Jackson's unreleased songs, he first estimated how many posthumous albums the Estate could release. He concluded from Jackson's past practice that each album would contain 11 songs. Since he believed Mijac held over 100 unreleased songs, he reasoned there would be 8-10 posthumous albums of original material after Jackson's death.

He then went to work to calculate the average revenue that each of these projected albums would earn. He began with the average sales of Jackson's last five original albums. He recognized the unlikelihood that Jackson's posthumous albums would be such blockbusters, and looked to other deceased singers' experiences to compute a ratio of posthumous-to-predeath sales. He concluded [*207] that the average sales for the first posthumous album would be 69.3% of average predeath album sales. He then projected the likely revenue per album domestically and internationally, not just for mechanical royalties, but also for artist royalties and joint-venture revenue.

For BMI royalties, Anson used what he thought were the annual averages from July 2004 through June 2009. Anson didn't project any revenue for synch fees.

Dahl. Dahl established base mechanical royalties for released Jackson songs and songs Jackson composed but didn't perform by averaging mechanical and other revenue from 2005 to 2008. Dahl noted that these numbers were likely inflated by Thriller 25's release in February 2008. He said he excluded all 2009 sales data because he believed that Thriller 25 made that year's data an outlier that wouldn't help him accurately predict long-term sales. When we looked closely at the numbers and not just the text of the report, however, we found that he does not seem to have used as his starting point the average of 2005 to 2008 sales but actual 2009 revenues.

For mechanical revenue for major and minor works by other artists, Dahl did a song-by-song projection using the average royalties that Mijac earned from them for the five years before Jackson's death. There is again what seems to be a [*208] disconnect between the text of his report and the data: The schedule that Dahl's text pointed us to shows data that goes back only to 2007. Then in his calculations, Dahl seemed to rely solely on the actual 2009 revenue results rather than any past average.

For mechanical revenue from unreleased songs, Dahl used the actual revenue Mijac earned from 2010. See infra p. 214. He did not state how many songs he estimated would be on posthumous albums.

For BMI and society rebates, Dahl stated that both Jackson's writer's share and Mijac's publisher's share need to be projected. He stated that he used a five-year average to determine a base income. Dahl does provide this. Once again in his calculations, however, Dahl used the actual results from 2009 as his base for projecting both BMI and society rebates. He also provided the historical data for only Jackson's writer's share, not Mijac's publisher's share. For synch fees, Dahl used actual synch fees that Mijac earned in 2009 — which were consistent with the fees it earned from 2004 through 2009 — $424,000.

This left us with a mess. Anson's credibility is already shaky at best, see supra pp. 58-60, and his report contains some errors on a cursory review. And Dahl's report has many inconsistences; he stated he was doing one thing but then in his calculations did something quite different. This inconsistency makes his [*209] calculations less reliable and his reasoning less persuasive. See Boltar, LLC v. Commissioner, 136 T.C. 326, 338-39 (2011); Estate of Bell v. Commissioner, 54 T.C.M. (CCH) 1123, 1125 (1987) (no confidence in expert report riddled with math errors). It also means that we ourselves need to figure out a starting point for each of the five groups of songs that Mijac held.

ii. Mechanical Revenue

Jackson's Compositions. For mechanical revenue from released Jackson songs as well as those that he composed but didn't perform, we first need to determine what dataset we should use. Dahl used data from 2005 through 2009. Anson didn't use data from 2005 and 2006. And the data that he did use was from an earlier draft report by Dahl that the Estate had turned over in 2014, which Anson didn't update with the final dataset that Dahl had assembled in 2016 from Mijac's records.

We will use Dahl's historical numbers that he used in his final report and that we admitted at trial. They are based on Mijac's own royalty statements from Warner and have had the benefit of more time to be proofread for accuracy. We find them to be the most reliable numbers we have.

We next need to decide whether to prorate royalties from the first half of 2009 to simulate a full year. Anson argued that a rational investor would use this [*210] number to project future earnings — specifically using the first half of 2009 to simulate a full year's worth of earnings. Dahl argued that projecting numbers from the first half of 2009 to the full year — which would show a 72% increase in revenue compared to 2007 — would allow a brief spike to unreasonably boost numbers for an entire year. Dahl would have us instead ignore 2009 all together.

We find the most reasonable approach to be in between. We won't ignore 2009 revenues, as they are the most recent data. We understand Dahl's position that these numbers may be inflated by the release of Thriller 25, but 2009 is one year out of five in the average — which allows the effect of Thriller 25 to be diluted. Odd and irrational spikes in revenue can occur with any artist. If a song is used in a popular ad or the soundtrack in a popular movie, or even if a singer's name pops up in the news, it is possible there will be a bump in sales. Such events seem to us to be stochastic; and if one uses five years of predeath data, it is likely to be smoothed out a bit. And by including the spike caused by Thriller 25's release, we find the data will better reflect the foreseeable, if somewhat unpredictable, spikes in revenue from the release of any other anniversary albums. We will not, however, project the spike for an entire year as Anson suggested. We find that extending this brief spike for an entire year would unreasonably inflate the value of Jackson's compositions. This means that we will include the first half of 2009 [*211] revenue in our average. We do this by breaking up the annual historical revenue in years 2005-2008 into half-year blocks.⁷⁸

We also need to distinguish between the two subgroups of Jackson compositions: released Jackson songs (which we find would be affected by a postdeath spike) and songs that he did not perform but for which he had a composer's credit (for which we find little evidence of any postdeath spike). The Estate provided this information, and we find it credible. We will take the average revenue from songs not performed by Jackson from 2005 through the first half of 2009. We therefore get two base revenues for Jackson's compositions: $3.6 million for released Jackson songs and $251,000 for songs that he composed but didn't perform. See infra Appendix B.

Major and Minor Works by Others. For mechanical revenue for both major and minor works by other artists, we will use the average revenue from 2007 through 2009. Dahl had these numbers in his final expert report. We will not project 2009 revenue by doubling the first half of that year's revenue, for the sake of consistency with how we projected revenue from Jackson's compositions (even [*212] though these numbers would not be affected by any postdeath spike). This leaves us with base royalties of $2.6 million for major works and $69,000 for minor works.⁷⁹ See infra Appendix C. We also need to determine which of these royalties are attributable to domestic sales and which are attributable to international sales, because (as we explain infra pp. 239-40) we think the growth rates for international and domestic sales are foreseeably different. Both Anson and Dahl agree that 41.1% of royalties would be generated internationally while 58.9% would be generated domestically. This creates a starting base for mechanical royalties for major and minor works of:

Source	Amount
Major works (domestic)	$1,547,965
Major works (international)	1,080,159
Minor works (domestic)	40,903
Minor works (international)	28,541

Songs. As a starting point for unreleased songs, we agree with both parties that the best way to project future cashflows is to project the value of posthumous albums with original songs. There were, of course, such posthumous albums whose sales are now known quite precisely. But we must figure out their value [*213] from information known (or reasonably foreseeable) as of the date of Jackson's death. Dahl doesn't estimate the average number of songs per album but didn't contest Anson's opinion that there would be 11 songs per album and we find that reasonable.

This brings us to another key question related to unreleased songs: How many of the unreleased songs are of high enough quality to be released? This is a valuation issue, but we do benefit from credible evidence that the Estate introduced. We must look at the marketability of these 83 songs. The Commissioner argues or assumes that most if not all of these songs could be released and sold, while the Estate argues that a great many are of poor quality.

We find that of these 83 unreleased songs, a reasonable buyer, as of June 2009, would believe that 22 were marketable. No reasonable buyer would believe that all 83 or even most of them were marketable, for several reasons. The most compelling reason for this finding is that Jackson was in a dire financial position for the last several years of his life and yet did not release any of these songs. We know that he was searching high and low — even going to usurious lenders — to bring in money. It is inconceivable that he wouldn't use unreleased songs to boost that income, and we find it more likely than not that he did not do so because he thought the vast majority were commercially nonviable; and we further reason that [*214] if Jackson himself thought this, so would hypothetical reasonable buyers and sellers. We therefore find that of the 83 songs available, a reasonably prudent buyer would believe only 22 were of high enough quality to be marketable. Twenty-two songs are enough to produce two posthumous albums.

We must next project sales of Jackson's first posthumous album from data that existed as of his death. We find Anson's projections reasonable as to this particular point, and we are satisfied with his method. We also note that Dahl offered no alternative since he relied only on actual sales data from 2010, which we must ignore as unavailable at the time of death.

We must determine the best number to use for average sales per album before Jackson died. Anson was reasonable in his calculations: He used only Jackson's last five albums, which correctly left out of his data an album such as Thriller, whose unprecedented success was anomalous not only in Jackson's career but also in the history of recorded music. The 2003 trial had an effect on Jackson's marketability; but as we've stressed throughout, Jackson's music was separate from the man. We will adopt Anson's average for Jackson's predeath domestic albums sold — 1.1 million.

Note that this is an average number of albums sold domestically. We think it reasonable to separately calculate foreign sales. Anson stated that according to [*215] his research, Jackson's revenue had historically been 44.7% domestic and 55.3% international. Dahl did not take issue with this, and neither will we. We can calculate Jackson's average predeath international album sales to be 1.3 million based on sales of Jackson's last five albums that he released before he died. We also adopt Anson's conclusion that a first posthumous album would have sales of 69.3% of the average predeath album. Dahl didn't dispute this either. This makes our projection of Jackson's first posthumous album sales: 747,802 domestic albums and 923,671 international albums.

The next question is how much revenue per album Mijac itself would receive. There was a big disagreement here: Anson based his calculations on what he thought Mijac would receive, not just as mechanical royalties but also artist royalties and joint-venture revenue from a deal between MJJ Ventures and Sony to exploit Jackson's master recordings. Dahl based his calculation on what he thought Mijac would receive as Jackson's share of the composer's share of mechanical royalties.

Dahl is correct — artist royalties and joint-venture income belong to the owner of the performances embodied in sound recordings. Mijac is entitled only to the royalties owed to a song's composer. This is so obvious a point that we suggested to Anson during trial that his inclusion of artist royalties and [*216] joint-venture revenue must be a mistake and he might want to recalculate his projections. Anson persisted in his error, but stubbornness receives no reward. Only composer royalties produce the revenue that Mijac would receive from unreleased songs. The parties stipulated that MJJ Productions collected Jackson's artist royalties under an agreement with Sony Music, Inc. for Jackson's services as a recording artist, and that MJJ Ventures collected Jackson's share of joint-venture income under an existing agreement with Sony for the exploitation of Jackson's master recordings. The parties negotiated a settlement of the value of these assets, and we won't let the revenue to which those ventures were entitled be counted again as future revenue for Mijac. Anson should have known this, as he did not include these other revenue streams in projection of revenue from Mijac from released Jackson songs or songs by other artists.

We can now focus on mechanical-royalty revenue and projections of income. For domestic albums, Anson projected composer's mechanical royalty revenue of 9.1 cents per song. Since the average album has eleven songs, this would mean on average $1.00 per album in mechanical royalties. Some of this revenue dribbles away in transaction costs (which includes Warner's administrative fees), and the experts agreed that from at least 1980 Jackson — or Mijac as the holder of Jackson's compositions — received 92.5% of this revenue.

[*217] Nothing else in the record suggests more or less in later years, so we'll round up and find that Mijac would net mechanical revenue of 93 cents per album sold domestically. Mechanical royalties are fixed by foreign law for international sales. Anson averaged four markets' mechanical royalty rates — Japan, Germany, France, and the UK — to come up with a projected mechanical royalty rate of 98 cents per album. Jackson himself had a contract with Warner to administer royalties from these sales, and under its terms, was to receive 75% of this foreign-income stream — which makes Mijac entitled to 73 cents per album sold internationally.

Dahl said that mechanical royalties need to be adjusted some more, because Jackson was not the only composer of all the songs he performed, and sometimes was not a composer at all. Dahl is without any doubt correct here. We have already found it reasonable to conclude that Jackson had only a 34% share of the unreleased songs. See supra p. 205. This means that Mijac would receive only 34% of mechanical-rights revenue from the unreleased songs.

In all, the base revenue for Mijac from Jackson's unreleased songs is $237,187 for domestic albums and $231,540 for international albums. See infra Appendix D.

[*218] Here is a summary of the starting point for our mechanical-revenue projections:

Mechanical revenue	Starting point
Released Jackson songs	$3,581,463
Jackson songs he didn't perform	250,826
Major works	Domestic: 1,547,965 International: 1,080,159
Minor works	Domestic: 40,903 International: 28,541
unreleased songs	Domestic: 237,187 International: 231,540

iii. Performance Revenue

To project future revenue streams from BMI and society rebates, we must be clear as to how these revenue streams earn money for Mijac and NHT III.

We start with BMI, the company that collected the composer's share of performance royalties for the compositions that Mijac owned as a publisher. BMI paid Warner — which administered Mijac — the publisher's share of this revenue. Warner would then pay Mijac its portion of royalties for the compositions that it owned.⁸⁰ Apart from Mijac, NHT III was also receiving royalties from BMI for [*219] Jackson's writer's share of performance royalties — a stream of income for NHT III that is separate from Mijac. BMI paid this income directly to Jackson — it didn't go through Warner. We must include it in our valuation of NHT III because, as the Estate concedes, Jackson had agreed to direct this money into NHT BI. to service its debt. And the loan documents with HSBC Bank confirm this — they state that part of the collateral for the loan is the compositions subject to the Warner agreement (i.e. Mijac) and the BMI agreement (i.e. Jackson's writer's share). We must, therefore, project Jackson's writer's share of performance income and Mijac's publisher's share of performance income.

We get little help from the experts here. Anson based his entire analysis on his interpretation of royalty data that was provided to him by Moss Adams rather than the actual royalty statements. Anson also incorrectly based his entire computation of all BMI performance revenue only on performance revenue Jackson himself received for his writer's share. This was, however, only one of the two sources of NHT III's income from BMI. And it wasn't even the source Mijac itself got. That makes his analysis of little help for projecting income from BMI.

[*220] Dahl understood this. He correctly stated that NHT III had two streams of revenue from BMI. He tried to project Jackson's writer's share of revenue. But he failed to, or at least did not adequately explain, his projection of revenue from BMI's payments for the publisher's share of the compositions Mijac owned.

The data is in the record, and we did what we could. We started with Jackson's writer's share of performance royalties from BMI. We have the quarterly reports from BMI to Jackson in the record.⁸¹ These show that from 2005 through the second quarter of 2009,⁸² Jackson's average annual revenue was just under $1.5 million. We need to distinguish between songs that Jackson composed and performed (which would be subject to a postdeath spike) and songs that Jackson composed but didn't perform. The statements do not distinguish these songs. We therefore find it most reasonable to apply the ratio that we do have from the data for mechanical royalties. That ratio was 93.45% for released Jackson songs and 6.55% for songs Jackson composed but didn't perform.

[*221] For the BMI revenue that went to Mijac through Warner, we looked to Warner's quarterly statements to Mijac. (These “statements” look less like bank statements and more like 10,000-row Excel spreadsheets.) These enabled us to identify income from BMI. From 2005 through the second quarter of 2009, Mijac earned roughly $1.2 million annually from BMI for its publisher's share. Of this, roughly 97.87% was domestic income.⁸³ See infra Appendix E.

This amount represents Mijac's publisher's share of performance royalties for all of the compositions it owns — not just Jackson's compositions. We need to further divide this revenue among the different groups of compositions that it holds. Once again we find it reasonable to infer that this breakdown is similar to mechanical revenue. Since we have data from 2007, 2008, and 2009 only for major and minor works, we will use the average revenue related to released Jackson songs from those three years. These calculations show that songs performed by Jackson accounted for 54.9% of revenue.

[*222] That leaves 45.1% of Mijac's revenue. We need to further break this down between songs that Jackson composed but did not perform, and works by other artists. This is because works by other artists are subject to recapture, while the songs written but not performed by Jackson are not. Based on mechanical revenue from 2007 to 2009 provided by Dahl, songs that Jackson composed but did not perform account for 6.03% of revenue, and other works account for 39.08% of revenue (38.07% for major works and 1.01% for minor works). See infra Appendix F. And performance royalties for major and minor works need to be further broken down to separate domestic revenue from international revenue because of a noticeable difference in the growth rate of BMI royalties depending on the place of performance. (We discuss this infra pp. 239-40.) We can get this from the Warner statements — 97.87% domestic and 2.13% international. See infra Appendix E. This creates the starting base for BMI revenue of:⁸⁴

[*223] Source		Amount
Writer's share directly from BMI:
Released Jackson songs		$1,391,121
Jackson songs he didn't perform		97,426
Publisher's Share from BMI through Warner:
Released Jackson songs		666,870
Jackson songs he didn't perform		73,225
Major works (domestic)		452,585
Major works (international)		9,827
Minor works (domestic)		11,959
Minor works (international)		260

For society rebates from BMI, we will use the average from 2004 to 2009. This dataset uses a June 30 year end, which eliminates the nagging question of whether to double-count data from the first half of 2009. (It does mean that the last two quarters of 2004 will be included, but the effect is minor.) This gives a starting base of $105,000 for society rebates for all of Mijac. But like BMI revenues, we need to break out society rebates attributable to each category. And we will also need to apportion for major and minor works and their domestic and international share. We will follow all the same percentages as with BMI. See infra Appendices E and F. With these assumptions, the starting base for society rebates is:

[*224] Source		Amount
Released Jackson songs		$57,861
Jackson songs he didn't perform		6,353
Major works (domestic)		39,269
Major works (International)		853
Minor works (domestic)		1,038
Minor works (domestic)		23

iv. Synch Fees

We adopt Dahl's uncontested opinion that the starting point for synch fees would be $424,000, which he based on a five-year average. We note that even this uncontested value is covered in a veil of confusion. Dahl states in his report that the base for projecting synch fees for released Jackson songs should be $424,000. He then states that because Mijac holds so many other compositions this projection should not exclude synch fees from these other groups. Seems reasonable. But once again in his actual calculations, Dahl seems to exclude any synch fees from the revenue projections of these other groups of compositions. Though something seems amiss, neither the Commissioner nor Anson challenges Dahl on projecting synch fees only for released Jackson songs. Neither shall we. And even if we wanted to and believed that these other works would likely [*225] produce some synch income, the record we have doesn't allow us any way to estimate how much.⁸⁵

Here's a summary of the starting point organized by source of revenue, and for mechanical-rights revenue further divided by asset group:

[*226] Source	Starting point
Mechanical revenue:
Released Jackson songs	$3,581,463
Jackson songs he didn't perform	250,826
Major works	Domestic: 1,547,965 International: 1,080,159
Minor works	Domestic: 40,903 International: 28,541
Unreleased songs	Domestic: 237,187 International: 231,540
BMI revenue:
Writer's share:
Released Jackson songs	1,391,121
Jackson songs he didn't perform	97,426
Publisher's share:
Released Jackson songs	666,870
Jackson songs he didn't perform	73,225
Major works	Domestic: 452,585 International: 9,827
Minor works	Domestic: 11,959 International: 260
Total society rebates	105,397
Synch fees	424,000

[*227] 3. The Spike/Projected Growth

The last — and what may be the most speculative — number on which the experts disagree is the magnitude of a postdeath spike. We are convinced by the experts' testimony and the data they presented that there is short-term spike in album sales (or nowadays, downloads) right after a musician of almost any degree of popularity dies. The height and duration of this spike likely reflects the extent of media coverage and his fans' shock at his death. There may also be something happening at a neurological level — scientific literature hints that the now well-known phenomenon that the frontal lobe is the last part of a person's brain to mature may somehow be connected with the common observation that people seem to stop liking new music at an unusually young age. See Carina Freitas et al., “Neural Correlates of Familiarity in Music Listening: A Systematic Review and a Neuroimaging Meta-Analysis,” 12 Frontiers in Neuroscience, Oct. 5, 2018 (meta-analysis of literature weakly suggesting emotional and hedonic responses activated in frontal lobe by familiar music). But see Arielle Bonneville-Roussy et al., “Music Through the Ages: Trends in Musical Engagement and Preferences From Adolescence Through Middle Adulthood,” 105 J. of Personality & Social Psychology 703, 703 (2013) (contending musical tastes do change over course of lifetime). Whether or not neuroscience and psychology ever figure this out, the [*228] experts in this case both agreed that something would propel a spike in demand for Jackson's music starting the moment his death was broadcast to the world. They disagreed sharply about how foreseeably high this spike would go, and how foreseeably long it would last.

i. Anson

Jackson-Released Songs. Anson relied on the average of seven comparable artists to estimate this postdeath spike. The average of these seven artists led him to predict that there would be a 461.6% increase in albums sales and performance royalties in the first year after death, a 47.2% decrease from this peak in the second year, a 15.6% decrease in the third year, a steady 10% decrease in years 4 through 6, and then a 5% decrease in years 7 through 10. He estimated that the phenomenon would then fade. For years 11 to 70 (the remaining life of Jackson's copyrights), Anson projected a decrease of 2.8% annually based on the compound annual growth rate in domestic recorded music revenues for the 15 years before Jackson's death. Anson assumed this spike would apply to all Jackson's compositions — both songs performed by Jackson and songs that were composed but not performed by Jackson.

The major problem with Anson's analysis is that six of the seven “comparable” artists with whom he compared Jackson weren't really comparable. [*229] The only one of the seven at all similar in popularity was Ray Charles, who had a sales spike of 1,529.3% in the first year after death, a 63.3% decrease in the second year after death, and a 53.9% decrease in the third year after death.⁸⁶

Major and Minor Works by Other Artists. Untouched by any postdeath spike would be Mijac's interests in copyrights in works by other artists. Anson expected these to maintain their historical average (2007-2009) for the first year after death and then decline 2.8% annually.

Anson assumed that these other artists would recapture their work at the earliest possible time. He determined an average recapture date of December 31, 2024 for major works and December 31, 2032 for minor works. After these compositions were recaptured, they would produce no further cashflow to Mijac.

Since recapture affected only his projections of domestic royalties, Anson needed to determine a useful life for international royalties. Rather than trying to project any future cashflow, Anson appears to have calculated only a terminal value for international royalties. This value for international royalties discounted to present value was $8.5 million. It is unclear to us how Anson calculated this [*230] value and why he chose not to project at least some future cashflows on an annual basis before calculating a terminal value.

Unreleased Songs. To estimate the posthumous sales of unreleased songs, Anson compared the sales of albums released after four comparable artists' deaths to their average album sales during their lifetime, and the first four years following those releases. As we've already discussed, the average revenue for a first posthumous album in the first year after a singer's death was 69.3% of his average sales on albums released while alive. Sales then decline 95.4% in the second year after release, 38.9% in the third year after release, and 30.2% in the fourth year after release. And, as difficult as it is to predict sales of a first posthumous album, it's even harder to predict sales of a second; of the singers that Anson analyzed, only two had more than one posthumous album. Average sales for second posthumous albums were 68.6% of the first, and sales of a third only 35.8% of the first. Following the first four years of an album's release, Anson assumes a decline of 2.8% annually. Since Anson believed Jackson had over 90 songs, he projected that a total of eight to ten albums would be released at a pace of one per year.

BMI Royalties and Society Rebates. Anson believed that the portion of BMI royalties related to Jackson's compositions would increase at the same rate of the [*231] posthumous spike of 461.2%, with a second-year decline of 47.2% and a third-year decline of 15.6%. The remaining BMI royalties (attributable to major works and minor works) would decline at a rate of 2.8% annually.

For society rebates, Anson projected a 2.8% decline in growth annually. He assumed no postdeath spike related to society rebates.

ii. Dahl

Dahl's numbers were drastically different. Dahl relied heavily on a 2011 research paper from the University of Zurich (Zurich Study) — which included the death of Jackson in its study and would not have been available at the time of death — to determine the amount of a postdeath spike. This study surveyed a wide variety of artists, including members of music groups (in contrast to solo artists).

Jackson-Released Songs. Based on this study, Dahl believed there would be a postdeath spike of only 50%. Because of the release of the 25th anniversary album of Thriller, Dahl incorporated a 20% growth in “base royalties” in year 1 and a 10% growth in year 2. After year 1, Dahl's computations for years 2-10 lack any sort of reasonable explanation for random spikes of growth in base royalties as well as other random numbers. (We guess that these random spikes might be attributable to potential album releases in the first 10 years after Jackson's death, [*232] but that'd be pure speculation.) Dahl predicted sales would shrink over the long term by 3% annually in years 11-70. Dahl did not explain this number.

Major and Minor Works by Other Artists. Dahl projected an annual growth rate in royalties for compositions from other artists — both minor and major works — of 1.7% for a 10-year period. This is one-half of the expected rate of growth in the music-publishing industry as reported by IBIS.⁸⁷ This low rate was due to the age of these songs. After 10 years, there will be an annual 3% decline in growth. Dahl, like Anson, projected that these songs would be recaptured at the earliest opportunity. For major works, Dahl employed a song-specific analysis, using each song's specific recapture year. For minor works, Dahl used a recapture year of 2029, and it is not clear to us where he got this.

Both experts assumed that there is no recapture date for foreign copyrights; so Dahl assumed that Mijac would earn royalties indefinitely from foreign sources. Dahl recognized that this was unreasonable, but argued nevertheless that this was a conservative approach, because it would overvalue Mijac's revenue stream.

[*233] Unreleased Works. For unreleased works, Dahl did not give a 10-year projection because he believed that the Estate would release albums with unreleased songs within the first 10 years after Jackson's death for fear that his popularity would rapidly decrease. We could not figure out how many albums Dahl believed would be released and when those releases would occur. On one page of his report, Dahl stated that he thought the Estate would release two albums soon after Jackson's death, based on the Estate's actual release of albums in 2010 and 2011, and then a new album in year 5, and another five years after that. In his actual calculations, however, Dahl assumed there would be two albums released in the first five years, two albums in the second five years, and then two more albums for the ten-year anniversary of Jackson's death. He said that this revenue projection for these albums combined actual posthumous sales and an assumption that the gradual decline in sales of Jackson's music during the last decade before death would continue. But when we looked at these projections for posthumous albums, we found that they matched actual revenue for 2010, 2011, 2013, 2014, and 2015, with differences only in 2012 and 2016 (accompanied by no explanation).

Dahl predicted that Jackson's posthumous compilation albums would, like Presley's, bring in only nominal revenue. (His report, however, also states that [*234] Presley's albums did reach the top 50 between 1977 and 1983, the years immediately after his death, but then sank, with no later album reaching the top 100 — even the album released to mark the tenth anniversary of his death.)

BMI and Society Rebates. We believe Dahl projected a two-year spike of 50% in this source of revenue with growth continuing up until five years after Jackson's death followed by a gradual decline — though this is far from clear from his report.⁸⁸ Dahl assumed that society rebates would follow BMI growth rates.

Synch Fees. Dahl also projected synch fees for Jackson-released songs over the next 70 years. He projected these would decline to the “steady state” that existed between 2004 and 2009, after a spike of 50% in year 1 and 25% in year 2, followed by a decline of 20% and 18% in years 3 and 4. In years 5 and 6, he predicted growth of 2.7%. In the next four years, Dahl projected growth of 1.5%, 1.1%, 0.7%, and 0% respectively. Over the last 60 years, the Estate projects a decline of 3% annually. This made for a roller coaster of spikes and declines:

[*235] Year	Spike/Decline percent
Year 1	50
Year 2	25
Year 3	(20)
Year 4	(18)
Year 5	2.7
Year 6	2.7
Year 7	1.5
Year 8	1.1
Year 9	.7
Year 10	0%
Years 11-70	(3) annually

Dahl gave little explanation for why he used these percentages. We note that Dahl did not value synch fees for any of the compositions other than released Jackson songs whose rights Mijac held.

iii. Our Opinion

Posthumous Spike. We begin with the postdeath spike, which affects a couple of these income streams. We find that — for the most part — Anson is more reasonable in his estimate. The Zurich Study that Dahl relied on was not available at the time of Jackson's death, and it included Jackson's actual postdeath numbers. It also included numerous lesser known musicians as well as data for values after [*236] the death of individual band members. Dahl failed to adequately explain his reasoning or methodology, and we even spotted several blatant mathematical errors. Just take a look:

• In projecting a spike related to Jackson-released songs, Dahl's 10-year projection has random growth spikes with no explanation.

• For BMI royalties, Dahl said he projected a two-year spike for performance revenue of 50% with growth continuing up until five years after death. The BMI projections don't show these spikes — they show instead a series of random spikes and declines with no explanation.

• In his projections of album sales, Dahl couldn't seem to stick to a single number. He at one point stated two albums will be released after death, with a new album in year 5 and another in year 10. In the actual calculations, Dahl stated he assumed there would be two albums released in the first five years, two in the second five years, and then two in year 10. And then his actual projections simply used revenues from Jackson's posthumous albums from years 2010, 2011, and 2013-2015.

This persistent divergence between his verbal explanations and his numbers undermine Dahl's overall credibility. See, e.g., Boltar, 136 T.C. at 338-39. We specifically find his conclusion that there would be only a 50% spike unreasonable. Most of Jackson's fans still considered him a great musician at the time of his death. Sales of his music seemed unaffected by the allegations of molestation — witness the near-instant sellout of shows immediately before his [*237] death. We need to look more closely at what a reasonably foreseeable postdeath spike would be.

Jackson's Compositions. We find Anson's projected growth rate over the 70-year life of Jackson's copyrights to be reasonable. We believe that at the time of death, a hypothetical buyer would reasonably have expected a potential spike for the King of Pop near that of Ray Charles. Anson, however, used a spike of only 461.6% — much lower than Charles's, but reasonable in that Charles lived a generation longer than Jackson. This means that his postdeath spike would more likely than not start from a lower base than Jackson's did. We will use Anson's postdeath spike of 461%.⁸⁹

To project years 4-10, we will also adopt Anson's projected decrease from that spike. We find that the underlying data Anson relied on is more reliable than that which Dahl used. It was also data that was available at the time of Jackson's death. We will likewise adopt Anson's forecast that after the first few years following Jackson's death, sales would start to shrink at an annual rate of 2.8%.

We do agree with Dahl that only released Jackson-performed songs would receive the postdeath spike. There was little, if any, evidence in the record that [*238] Jackson's popularity as a performer flowed from his own compositions — his albums and concerts included a great many performances where he sang the works of others. Since we calculated our base revenue for Jackson compositions; $3.6 million for released Jackson songs and $250,826 for songs Jackson composed but didn't perform — we have the data to apply a postdeath spike only to the songs performed by Jackson. The songs composed by Jackson but not performed by him will be treated just like songs written by the other artists that are included in Mijac's assets. See infra pp. 238-40.

This brings total projected mechanical revenue related to released Jackson songs, discounted to present value, to be $78.4 million.⁹⁰ And the total projected mechanical revenue related to Jackson's compositions that he didn't perform discounted to present value, to be $2.7 million. See infra Appendix G.

Major and Minor Works by Other Artists. For projected mechanical revenue from both major and minor works by other artists, we find merit in parts of each expert's projections. We do find it reasonable that revenue from these works would increase somewhat over the next 10 years, but we find Dahl's expected growth rate — 1.7% — to be more reasonable. Copyrights are predictably [*239] wasting assets, however, and we will use Anson's long-term trend of revenues shrinking 2.8% after that.

We agree with both parties that these songs are likely to be recaptured at the first opportunity. For major works we will use the recapture date of December 31, 2024. We believe that using an average recapture year of 2024 is more than adequate. Since Dahl did not explain how he determined his projected recapture year for the minor works, we will use Anson's date of December 31, 2032. Cashflow projections for domestic royalties should be projected only to 2024 for major works and 2032 for minor works.

We are unsatisfied with both parties' projections of international royalties. We don't understand why Anson chose not to calculate any projected cashflows — despite using the DCF analysis — and instead calculating only a terminal value. And though we appreciate Dahl's candor, we are reluctant to project international royalties indefinitely — something that Dahl admitted was unreasonable.

We believe the best way forward is to project international cashflow to 2024 for major works and 2032 for minor works.⁹¹ Both Anson and Dahl agree that 41.1% of royalties are generated internationally and 58.9% are generated [*240] domestically, and we will also use these percentages. After this, we will calculate a terminal value using the Gordon Growth Formula. With these assumptions, the projected mechanical revenue of major works discounted to present value is $31.8 million. See infra Appendix H. The projected mechanical revenue for minor works discounted to present value is $882,812. See infra Appendix I.

Unreleased Songs. For posthumous albums, we were dissatisfied with both experts. Dahl stated three different album projections over the ten years following Jackson's death. And his calculations show random spikes in growth with no explanation of how he determined them.

We find Anson's projection on this point to be more reasonable. Though Anson's sample size was limited, we find his prediction of average sales more reliable than what seems to be the random projections that Dahl made. Dahl also admitted that he based his projections on the actual results of posthumous released albums containing Jackson's unreleased works.

Since we believe that a reasonable investor would believe that there were 22 marketable unreleased songs at Jackson's death, using the average of 11 unreleased songs per album, we believe there would only be two albums' worth of posthumously released songs. We need not take into consideration here the [*241] release of anniversary albums, as the effect of these is already included in our projections with Jackson-released works since we included the Thriller 25 spike.

For the growth rate per album, we are persuaded to use the growth rate used by Anson. For the first posthumous album, we already established that the starting point of the projections will be 69.3% of the average number of albums sold. See supra p. 215. In year 2, sales will decrease 95.4%, in year 3 38.9%, and in year 4 30.2%.

For the second posthumous released album, we also find Anson's analysis reasonable, and will project revenue from a second album of 68.6% of sales from Jackson's first posthumous album, spread over the first four years after its release. To produce year-by-year projections, we will simply decrease the amounts for the second album by 31.4% from the first album's projections.

After four years, we forecast that revenue from each album will decrease at a rate of 2.8%. Anson projected cashflows for 16 years after Jackson's death, while Dahl projected cashflows for only 10 years. We believe that it is reasonable to project 10 years and then calculate a terminal value for the remaining life of the asset (years 11-70) using the Gordon Growth Formula. [*242] With these assumptions, we calculate that the value of mechanical royalties from Jackson's unreleased songs discounted to present value was $961,667 at the time of death. See infra Appendix J.

We can summarize all the projected mechanical revenue discounted to present value:

Source	Amount
Released Jackson songs	$78,372,575
Jackson songs he didn't perform	2,730,827
Major works	31,763,802
Minor works	882,812
Unreleased songs	961,667

BMI Royalties and Society Rebates. Dahl's lack of clarity makes it difficult to believe his projections. We do agree with both parties that BMI royalties should mirror the growth rates of mechanical revenue for released Jackson songs, songs Jackson composed but didn't perform, major works, and minor works. It seems reasonable to project that, just as Jackson's posthumous popularity spiked after his death, his songs would be aired more in the media that BMI monitors — and roughly mirror the growth rates for Jackson-released songs that we've already discussed. Royalties from songs not associated with Jackson in the popular mind should not be affected by his death, and we therefore find they [*243] should have the same growth rate as before he died. We therefore find it reasonable to adopt the same projection rate for BMI royalties as released Jackson songs, songs Jackson composed but didn't perform, major works, and minor works. See infra Appendices K, L, and M.

Though this seems simple, this mirroring requires digging through the BMI revenue and allocating it to its source — which we already did to determine our starting base. See supra pp. 220-21.

Songs that Jackson composed but did not perform will mirror the normal growth rate from major and minor works, increasing at 1.7% for 10 years then decreasing 2.8% for their remaining life. Other works will still be subject to recapture, which again will affect only domestic royalties. We will use the same recapture dates that we discussed above — 2024 for major works and 2032 for minor works. See infra Appendix M.

With these assumptions, we can project total cashflow for BMI (not discounted to present value). Here is the breakdown:

[*244] Source	Amount
Writer's share:
Released Jackson songs	$91,560,188
Songs Jackson didn't perform	4,344,842
Publisher's share:
Released Jackson songs	43,891,738
Songs Jackson didn't perform	3,265,571
Major works (domestic)	7,433,014
Major works (international)	161,396
Minor works (domestic)	283,049
Minor works (international)	6,146

When we discount them to present value we get $51.2 million — $31.5 million for Jackson's writer's share and $19.7 million for Mijac's publisher's revenue. See infra Appendices K and N.

Society rebates were sporadic before Jackson's death, and we'd expect them to remain so after his death. Dahl simply asserted that growth of an entire catalog will follow BMI royalties related to Jackson's mechanical royalties, but we can't see how that would be so. We find that Anson's projections are more reliable, because he at least acknowledged the problem of valuing a catalog-wide fee and didn't include a postdeath spike. We find it reasonable to forecast an annual decline of 2.8%.

[*245] As with BMI royalties, we do need to break out the society rebates attributable to Jackson-released songs, Jackson songs he composed but didn't perform, major works, and minor works (which also will be broken out domestically and internationally). We will follow all the same percentages that we used to project BMI royalties. See infra Appendix O and P. With these assumptions here is the breakdown of future cashflows (not discounted to present value):

Source	Amount
Released Jackson songs	$1,733,490
Jackson songs he didn't perform	190,345
Major works (domestic)	472,862
Major works (international)	10,267
Minor works (domestic)	17,276
Minor works (international)	375

The total value of society rebates discounted to present value is $808,943. See infra Appendix Q.

Synch Fees. Anson gave us no analysis or projection for synch fees. Dahl gave us numbers, but a contradictory explanation. Dahl stated that the growth rate for synch fees would follow the growth rate of released Jackson songs. We find this reasonable and, therefore, find it reasonable to use a growth rate for synch fees [*246] equal to the growth rate of released Jackson songs. Synch fees discounted to present value are $9.3 million. See infra Appendix R.

We can add all the revenue streams and summarize here:

Income stream	Present value
Mechanical:
Released Jackson songs	$78,372,575
Jackson songs he didn't perform	2,730,827
Major works	31,763,802
Minor works	882,812
Unreleased songs	961,667
BMI:
Publisher's share	19,664,865
Society rebates	808,943
Synch fees	9,278,324
Total value of Mijac	144,463,816
BMI writer's share	31,502,394
Total value of all revenue streams	175,966,210

D. NHT III

Since the rest of NHT III's assets and liabilities were not in dispute we can quickly calculate its value:

[*247] Assets
Cash	$3,500,000
Mijac Music & writer's share	175,966,210
Total assets	179,466,210
Total liabilities	(72,152,649)
FMV	107,313,561

IX. Penalties

The last issue for us to resolve is whether the Estate is liable for penalties. The Commissioner's determination of penalties flowed in part from disputes about the value of Jackson's assets other than the three we had to analyze in this opinion. The parties agreed before trial that penalties wouldn't apply to those settled issues. We are left to decide whether any accuracy-related penalties apply to the Estate's reported values of three assets at issue.

We first consider whether the Commissioner complied with section 6751. That section states that no penalty is allowed unless the “initial determination of such assessment is personally approved (in writing) by the immediate supervisor of the individual making such determination.” Sec. 6751(b)(1). This written approval must be obtained no later than the date the notice of deficiency is issued or the date the Commissioner files an answer or amended answer in which he asserted the penalty. Chai v. Commissioner, 851 F.3d 190, 221 (2d Cir. 2017), [*248] aff'g in part, rev'g in part 109 T.C.M. (CCH) 1206; see also Graev v. Commissioner, 149 T.C. 485, 493 (2017), supplementing and overruling in part 147 T.C. 460 (2016).

Section 7491(c) places the burden of production on the Commissioner with respect to the liability for any individual for any penalty. Since an estate is not an individual, Estate of Ramirez v. Commissioner, 116 T.C.M. (CCH) 517, 524-25 (2018), the Estate bears the burden of production here. This is a somewhat unusual burden — the burden of producing evidence that no evidence exists of the Commissioner's compliance with his obligation to show supervisory approval of penalties. We described the problem at greater length in our order denying the Commissioner's motion to reopen the record. See Order dated December 20, 2017. But the Estate failed to enter any evidence that the Commissioner didn't comply with section 6751.

When the burden lies on the Commissioner, we have held that he must introduce evidence that he complied with section 6751 to meet his burden of production. See, e.g., Dynamo Holdings Ltd. P'ship v. Commissioner, 150 T.C. 224, 227 (2018) (stating that when the burden of production lies with the Commissioner, he must enter evidence of supervisory approval); Higbee v. Commissioner, 116 T.C. 438, 446 (2001) (“[F]or the Commissioner to meet his [*249] burden of production, the Commissioner must come forward with sufficient evidence indicating that it is appropriate to impose the relevant penalty”). We see no reason the Estate shouldn't be held to the same standard.

That means the Estate could have entered into the record its discovery request for any written supervisory approval or other evidence to show that it had raised this issue. See Lew v. Moss, 797 F.2d 747, 751 (9th Cir. 1986) (“[T]he burden of production has the sole effect of forcing * * * [its holder] to produce enough evidence to avoid a direct verdict”); Jordan v. Herrera, 224 F. App'x 657, 658 (9th Cir. 2007); see also Gathright-Dietrich v. Atlanta Landmarks, Inc., 452 F.3d 1269, 1274 (11th Cir. 2006). The Estate could even have raised the issue in the list of assignment of errors in its petition. See Wheeler v. Commissioner, 127 T.C. 200, 206-07 (2006), aff'd, 521 F.3d 1289 (10th Cir. 2008). Thriller is part of the record here. So are demons, vampires, monsters, ghosts, and even the funk of 40,000 years. But the record lacks any evidence that the Commissioner's agent failed to obtain supervisory approval.

We shift to the merits of the penalty. The Commissioner asserts the Estate is liable for several section 6662(a) penalties: gross-valuation misstatement under section 6662(h), substantial-valuation misstatement under section 6662(g), and negligence or disregard of the rules under section 6662(c). Since we found that [*250] the Estate correctly valued NHT II, no penalties will apply to any underpayment due to its valuation. This leaves the valuations of Jackson's image and likeness and NHT III.

These penalties can be negated by proof that the Estate had reasonable cause and good faith for its return position. Sec. 6664(c). We determine whether this defense applies on a case-by-case basis, after looking at all pertinent facts and circumstances. Sec. 1.6664-4(b)(1), Income Tax Regs. The most important factor is the taxpayer's effort to assess his liability. Id. Just because there was an appraisal of the asset in question does not per se show reasonable cause and good faith. Id. Instead, we must look at the appraisers' assumptions, the appraised value, and the circumstances under which it was obtained. Id.

The Estate valued Jackson's image and likeness at roughly $2,000 on its return. This was based on an appraisal by Moss Adams — an accounting firm that we specifically find is reputable and credible. See supra pp. 49, 100-01. Moss Adams based its valuation on the last 10 years of Jackson's life, and concluded to the surprise of the Estate that Jackson's image and likeness was not worth very much. The Commissioner argues that this valuation was clearly wrong — and the Estate should have known not to rely on it.

[*251] But the facts show that this low valuation wasn't that farfetched. Jackson made almost no money attributable to his name and likeness in the last decade of his life, especially after the 2003 trial. And in 2009, even as Jackson rapidly sold out multiple concerts, exploitation of his name and likeness earned him only $24. Moss Adams followed standard appraisal procedure in this area — it focused on the last 10 years of Jackson's life. Though Fishman and Roesler — who we find credible — eventually expanded their dataset, they both stated in their reports that they typically only look at the 10 most recent years of income. While we disagree with Moss Adams's appraisal, we do find that it was reasonable. And we find that the Estate reasonably relied on it in good faith once it discovered how little revenue Jackson had been earning from use of his name and likeness. No penalties here.

We find much the same for the valuation of NHT III. Our own opinion shows how complicated valuing Mijac and NHT III is. The Estate again used Moss Adams to help them estimate this value. We also disagree with this appraisal value, but we again find that it was reasonable given all the facts and circumstances. And we again find it was reasonable for the Estate to rely on it and that it did so in good faith. No penalties here either.

[*252] X. Conclusion

Jackson had outlived the peak of his popularity, but in the decades before his death he kept spending as if he had not. Plausible allegations of repellent behavior ruined his personal reputation, and with it his ability to earn much income apart from his music. He went deeply into debt to keep his life as it had been. Those debts increased; the interest on them rose; bankruptcy was a foreseeable outcome.

These troubles affect our factfinding. We have to look for the value of each of Jackson's assets as if “in the decedent's hands at the time of its transfer by death.” Estate of Simplot, 249 F.3d at 1194-95. The value we put them as of the day he died is, we acknowledge, much less than their value much later under the Estate's management. Branca, a friend of Jackson's for many years, but a practical man forever, credibly testified that as popular singers age their prominence declines. Jackson, at the time of his death, was not behaving as if this were true; even a rational and undistressed hypothetical seller would have been hardpressed to avoid fire-sale prices.

But Branca is right. Older stars' “fans are less apt to buy tchotchkes.” Older stars do get less play on the radio. And according to all the expert witnesses, the same is true about even Jackson. They predicted that, like most pop [*253] stars, Jackson would have a foreseeable surge in sales of his songs when he died, but a surge that would fade with time.

They are right. Popular culture always moves on. There will come a time when Captain EO joins Monte Brewster and Terry Forbes as names that without googling sort of sound familiar, but only to people of a certain age or to students of entertainment history. And just as the grave will swallow Jackson's fame, time will erode the Estate's income. It resurrected and then sold what became its most valuable asset to Sony before trial. The value of what it has left, no matter how well managed, will now dwindle as Jackson's copyrights expire and his image and likeness shuffle first into irrelevance and then into the public domain.

Before any such general reckoning, however, we must have more exact computations. We can summarize our findings on the three issues that remain:

Asset	Estate	Commissioner	Tax Court
Jackson's image and likeness	$3,078,000	$161,307,045	$4,153,912
NHT II	-0-	206,295,934	-0-
NHT III	2,267,316	114,263,615	107,313,561

Decision will be entered under Rule 155.