In Re THP Capstar Acquisition Corp.

756 F.Supp.2d 516 (2010) In re Application of THP CAPSTAR ACQUISITION CORP. (now known as DMX, Inc.). Related to United States of America, Plaintiff, v. American Society of Composers, Authors, and Publishers, Defendant. Nos. 09 Civ. 7069(DLC), 41 Civ. 1395(DLC). United States District Court, S.D. New York. December 1, 2010. *517 Christopher J. Glancy, Stefan Mentzer, Boris Sharapan, White & Case LLP, and Richard H. Reimer, Samuel Mosenkis, New York, NY, for ASCAP. R. Bruce Rich, Benjamin E. Marks, Todd D. Larson, Weil, Gotshal & Manges LLP, New York, NY, for applicant DMX, Inc. OPINION AND ORDER DENISE COTE, District Judge. Table of Contents INTRODUCTION ........................................................... 518 FINDINGS OF FACT ....................................................... 520 *518 A. ASCAP ........................................................... 520 B. DMX ............................................................. 520 C. ASCAP's Blanket Licenses with the Background/Foreground Music Industry ...................................................... 522 1. The 1987 and 1994 Form License Agreements .................... 522 2. Post-1999 Litigation Over a Per-Segment License .............. 523 3. 2005 Muzak Agreement ......................................... 524 4. 2010 Music Choice Agreement .................................. 525 5. 2009 PlayNetwork Agreement ................................... 526 6. 2009 Trusonic Memorandum of Understanding .................... 527 D. Competitive Environment in the Background/Foreground Music Services Industry ............................................. 527 E. DMX's Direct License Program .................................... 528 1. Key Terms of the MCCL ........................................ 529 2. The MCCL Program from 2006 to 2010 ........................... 532 3. Renewals of MCCLs ............................................ 532 4. Refusals to Renew MCCLs ...................................... 533 5. DMX's Outreach to "Major" Publishers The Sony MCCL ........... 533 6. DMX's Negotiations with Other Major Music Publishers ......... 535 F. DMX's Application for an ASCAP Blanket License .................. 535 G. DMX'S Application for a BMI Blanket License ..................... 535 CONCLUSIONS OF LAW ..................................................... 536 A. The Governing Law ............................................... 536 B. ASCAP's Rate Proposals .......................................... 539 1. ASCAP's First Proposal A Flat Fee Blanket License ............ 539 2. ASCAP's Alternative Proposal A Blanket License with a Static Carve-Out .................................................. 541 C. DMX's Proposal .................................................. 547 1. The Unbundled Music Fee ...................................... 548 2. The Floor Fee ................................................ 548 3. The Blanket Fee Sum of the Floor Fee and the Unbundled Music Fee ........................................................ 549 4. The Structure of the DMX Proposal is Reasonable and the Benchmarks that DMX Proposes are Appropriate ............... 549 CONCLUSION ............................................................. 552 INTRODUCTION DMX, Inc. (formerly known as THP Capstar Acquisition Corp.) ("DMX"), a leading background and foreground ("BG/FG") music service provider, has asked the American Society of Composers, Authors and Publishers ("ASCAP"), a membership organization representing almost half of American composers and music publishers in their negotiations of public performance rights, for a license. DMX seeks a through-to-the-audience blanket license adjusted to reflect the many licenses DMX has already obtained directly from music publishers and the many more it intends to obtain. The parties have been unable to reach agreement on the terms of such a license, and, pursuant to an antitrust consent judgment, ASCAP now requests that this Court set a rate for the license. The parties agree that, at least in theory, such a license — a blanket license with "carve-outs" for DMX's direct licensing program — is permitted by the consent judgment under which ASCAP operates. They dispute, however, the extent to which such a license is "reasonable" and whether ASCAP may reasonably respond to a request for a blanket license with carve-outs with a proposal for a blanket license that does not account for a music user's direct *519 license agreements with ASCAP publishers. Additionally, the parties' dispute what the appropriate benchmark agreements are for determining a reasonable license fee, and what rate is indicated by those past agreements. On July 25, 2006, ASCAP applied to this Court to set a reasonable rate.[1] A bench trial was held from November 15 to 23, 2010, to determine a reasonable rate pursuant to DMX's application to ASCAP for a license and ASCAP's application to this Court. This Opinion constitutes the Court's findings of fact and conclusions of law following that trial. The factual findings are principally set forth in the first section of this Opinion, but appear as well in the final section. With the parties' consent, the trial was conducted in accordance with the Court's Individual Practices. On September 10, the direct testimony of the witnesses was presented through affidavit and submitted with the joint pretrial order, along with the parties' trial exhibits and proposed findings of fact and conclusions of law. ASCAP presented affidavits constituting the direct testimony of two of its employees and one expert. Its employee-witnesses were Vincent Candilora, Director of Licensing, and Dr. Peter M. Boyle, Chief Economist ("Boyle"). ASCAP's expert was Dr. William H. Greene ("Greene"),[2] an economist. DMX presented affidavits constituting the direct testimony of three of its employees, a consultant, and two experts. The DMX employees were Timothy J. Seaton, Chief Operating Officer ("Seaton"), L. Barry Knittel, Senior Vice President of Business Affairs Worldwide ("Knittel"), and Shalonn Hilburn, Manager of Select Music Design. DMX's consultant was Ronald H. Gertz ("Gertz"), Chairman of Music Reports, Inc. ("MRI"). DMX's experts were Dr. Adam B. Jaffe,[3] an economist ("Jaffe"), and Dr. Amy Candell,[4] a consulting economist ("Candell"). All of the parties' witnesses *520 appeared at trial and were available for cross-examination. In addition, the parties designated deposition testimony from eleven witnesses. The additional witnesses for whom ASCAP offered deposition testimony were Edward Arrow, Vice President of Copyright at Universal Music Publishing Group ("Universal"); Helene Blue, President and sole full-time employee of Helene Blue Music; Philip Ciadella, Senior Vice President of Administration and Licensing at Cherry Lane Music Publishing ("Cherry Lane"); Alan Furst, Senior Vice President of Content for DMX; Benjamin Hanson, currently an attorney for Harden Healthcare Services and General Counsel for DMX Holdings, Inc. from June 2005 until September 2008; David Hirshland, President of Bug Music, Inc.; Jonas Kant, Senior Vice President of Business and Legal Affairs for Sony/ATV Music Publishing ("Sony"); Sindee Levin ("Levin"), music publisher and owner of the American Mechanical Rights Agency, Inc.; and Brian Roberts, Chief Financial Officer for Warner/Chappell Music Publishing ("Warner/Chappel"). The additional witnesses for whom DMX offered deposition testimony were John LoFrumento, ASCAP's Chief Executive Officer; Robert Candela, ASCAP's Chief Financial Officer; and Richard Reimer, ASCAP's Senior Vice President of Legal Services. FINDINGS OF FACT A. ASCAP Created in 1914, ASCAP is an unincorporated membership and performing rights organization ("PRO") controlled by composers and music publishers. ASCAP currently has approximately 400,000 United States members. ASCAP members grant ASCAP a nonexclusive right to license the public performance right to their musical compositions. There are millions of musical compositions in ASCAP's unique repertory. ASCAP licenses its members' music to a wide variety of music users, including BG/FG music services.[5] ASCAP's ability to collectively license the performing rights to its members' compositions gives the organization market power when negotiating with music users. In an effort to police ASCAP's exercise of this power, the Department of Justice ("DOJ") sued ASCAP in 1941 for alleged violations of the Sherman Antitrust Act. This lawsuit resulted in a consent decree that has constrained ASCAP's operations. United States v. ASCAP, 1940-43 Trade Cas. (CCH) ¶ 56,104, § II(1) (S.D.N.Y. 1941). The ASCAP consent decree has been updated on two occasions. United States v. ASCAP, 1950 Trade Cas. (CCH), ¶ 62,595 (S.D.N.Y.1950) ("Amended Final Judgment" or "AFJ"); United States v. ASCAP, 2001-2 Trade Cases (CCH) ¶ 73,474 (S.D.N.Y.2001) ("AFJ2"). The most recent version of the decree — AFJ2 — took effect on June 11, 2001. AFJ2, 2001-2 Trade Cases (CCH) ¶ 73,474. B. DMX DMX is one of five leading competitors in the BG/FG music industry, each of which services more than 10,000 locations.[6] Generally, BG/FG music service businesses *521 provide their clients with pre-programmed music for use in a variety of business establishments, from restaurants and hotels to offices and retail stores. The right to publicly perform musical works is a part of the service that BG/FG companies provide to clients. DMX was formed June 3, 2005, when DMX Holdings, Inc. purchased assets out of the bankruptcy estate of Maxide Acquisition, Inc. and its subsidiaries. Approximately half of the customer locations serviced by DMX are associated with businesses that control more than 100 locations. DMX's customers vary in size, from national chains like Saks Fifth Avenue, Sheraton, and Olive Garden to single-location businesses. In 2005, DMX served approximately 84,000 locations. This number had dropped to 70,000 locations by late 2009, but recently rose to approximately 95,000 after DMX entered into an agreement with DIRECTV, Inc. ("DIRECTV") to take over the audio channels that are provided as part of DIRECTV's commercial satellite television service. In any given year, DMX delivers approximately 150,000 musical compositions to its customers. These works are owned or controlled by more than 14,000 different publishers. From 2005 to 2009, ASCAP works constituted approximately 48% of all of the musical works that DMX transmitted or delivered to its clients. DMX provides music programming to its customers through either an "off-premise" or "on-premise" delivery mechanism. Off-premise delivery involves the transmission of music to customers via direct broadcast satellite ("DBS"). A satellite dish installed at each business establishment is connected to a proprietary DMX receiver that receives the broadcast signal and transmits it through the establishment's sound system. Businesses that subscribe to DMX through its "on-premise" delivery service are given a proprietary DMX playback device that is installed at the customer location. This device is updated either through the delivery of physical media, such as a CD or DVD, or through a network connection. DMX's customers are fairly evenly divided between DBS and on-premise delivery mechanisms. On-premise and off-premise DMX customers have access to a large array of programming. The off-premise service consists of 111 different audio channels that vary in terms of genre and style. Approximately half of the 52,000 off-premise customers receive all of the available channels. The other half only receives 84 channels as part of DMX's delivery to commercial establishments through DIRECTV. A small number of customers receive packages of two, six, or ten channels. DMX does not track which channels its off-premise customers actually play,[7] but it does keep a list of every song that it broadcasts and the number of times that any given song is transmitted on each channel in any given period. On-premise customers have access to 157 programs.[8] The on-premise and off-premise *522 versions of the same program have between 90% and 95% of the tracks in common. But, because recorded music must be physically delivered to on-premise establishments, these customers are required to select a smaller number of programs—usually ten—for physical delivery. Customers receive updated playlists for their selected programs every thirty to ninety days. DMX tracks which songs are included on each program's playlist and the number of locations that request each program. It does not keep records of how often customers play any given program or how often they repeat a particular song. DMX provides two categories of services to its customers: Select and Signature. Select customers receive all or some number of the various channels discussed above—these channels are organized by category, i.e. by genre, era, or a particular "lifestyle." All Select customers are given the same playlist, either via DBS or some form of physical media. Signature customers, however, receive programs that are specifically designed for them by DMX personnel. These programs are delivered almost exclusively via on-premise media. In 2008, approximately 30% of DMX's customers received the Signature service, but this number has since declined. C. ASCAP's Blanket Licenses with the Background/Foreground Music Industry 1. The 1987 and 1994 Form License Agreements Prior to 2000, ASCAP licensed the entire BG/FG music industry under two form licenses. From January 1, 1987 through December 31, 1991, ASCAP offered a form license (the "1987 Form License") to publicly perform ASCAP works by means of a BG/FG music service at the premise of the licensee's subscribers. The 1987 Form License distinguished among three types of businesses that use BG/FG music services. The annual fee for a license to provide BG/FG music to retail businesses-known as "5A" locations— ranged from $38.50 per location in 1987 to $44.50 per location in 1991. Licenses for non-retail locations, such as offices, hospitals, or universities, ranged from 4.5% to 5% of the BG/FG music service's "gross billings" for these customers.[9] Finally, for establishments that use "commercial announcements" in addition to the BG/FG music service, the license went from an annual rate of $50.40 "per floor" in 1987 to $59.40 per floor in 1991. The latter two categories were known as "5B" and "5C" locations, respectively. After the 1987 Form Agreement expired on January 1, 1992, the major BG/FG music service providers continued to be licensed by ASCAP and paid, on an interim basis, the rate applicable to them in December 1991 under the 1987 Form Agreement. From June 1, 1994 until May 31, 1999, ASCAP offered a second form agreement (the "1994 Form Agreement") to BG/FG music service providers. All major BG/FG music service providers signed this agreement, including the two companies that became modern-day DMX. This agreement paralleled the 1987 Form Agreement in structure, but provided for an increase in license rates: the annual rate for 5A retail *523 establishments started at $46.25 per location in 1994 and rose to $49.50 per establishment by 1999; a license for 5B non-retail locations cost 5% of gross billings for each location; and, the annual cost of a license for a 5C establishment with "commercial announcements" began at $61.68 per floor in 1994 and climbed to $66.00 per floor in 1999. When the 1994 Form Agreement expired on June 1, 1999, all major BG/FG music service providers continued to operate under an ASCAP blanket license and paid interim fees at the rates that were applicable in May 1999. 2. Post-1999 Litigation Over a Per-Segment License Efforts to reach a third form license agreement with leading members of the BG/FG music industry failed. In early 2003, ASCAP filed an application with the rate court for a determination of a reasonable fee as to the predecessor in interest to DMX—DMX Music, Inc.—and one if its principal competitors, Muzak. United States v. ASCAP (In re Application of Muzak, LLC), 309 F.Supp.2d 566, 570 (S.D.N.Y.2004) ("Muzak I"). During that litigation, DMX Music, Inc. and Muzak indicated their desire to obtain a "per-segment" license from ASCAP, with each segment defined as a particular music publisher's catalog.[10]Id. They sought to have their payments to ASCAP measured by the "degree to which [they] rely on an ASCAP blanket license, as opposed to direct licensing arrangements with ASCAP members." Id. ASCAP refused to quote a fee for such a license, arguing that it was a form of license that it was not obligated to provide under AFJ2. Id. at 571. In March 2004, the rate court held that a music catalog "does not involve the user's performance of music and therefore is not a `segment' under the relevant clauses" of AFJ2. Id. at 572. Muzak I went on to hold, however, that "the `per segment' licens[e] provision[]" of AFJ2 "do[es] not preclude the issuance of a blanket license with a fee structure that reflects applicants' previous direct licensing arrangements." Id. at 580. The Honorable William C. Conner, who previously served as the ASCAP rate court judge, concluded that the Second Circuit's decision in United States v. BMI (In re Application of AEI Music Network), 275 F.3d 168 (2d Cir.2001) ("AEI"), which interpreted the consent decree under which ASCAP's principal competitor BMI operates, did not "preclude the issuance of blanket licenses with reasonable fees that reflect direct licensing arrangements previously entered into by applicants." Muzak I, 309 F.Supp.2d at 578. Since the operative provisions in the BMI consent decree were identical to provisions in AFJ2, Judge Conner opined that the existence of ... direct licensing arrangements may and will be considered by this Court in a rate court proceeding under AFJ2 section IX in determining whether ASCAP has met its burden of proving the reasonableness of the blanket license fee it seeks or, in the event that ASCAP fails to meet that burden, in the Court's calculation of a reasonable fee based on all the evidence. Id. at 581 (emphasis supplied). None of the parties to the proceedings before Judge Conner appealed his decision. A few months later, in July 2004, Judge Conner clarified that Muzak I did not "contemplate a blanket license fee mechanism that provides credits or discounts for direct licensing arrangements that applicants may enter." United States v. ASCAP (In re Application of Muzak, LLC), 323 F.Supp.2d 588, 590 (S.D.N.Y.2004) *524 ("Muzak II"). Instead, the rate court was only required to consider those direct licenses that were "already in existence at the time of trial in determining a reasonable blanket licensing fee." Id. In the aftermath of that rate court litigation, ASCAP and Muzak settled their disputes and entered into a licensing agreement. DMX Music, Inc. filed for bankruptcy. During the proceedings that produced the Muzak I and Muzak II decisions, ASCAP had signed letter agreements with two other major DMX competitors—Music Choice and PlayNetwork—in which the parties agreed to interim fees and ASCAP promised to offer them a license on the same terms ultimately given to Muzak and DMX Music, Inc. 3. 2005 Muzak Agreement ASCAP relies principally on its 2005 agreement with Muzak, concluded in the wake of the Muzak rate court proceedings, as the most appropriate benchmark for constructing a blanket license for DMX. Muzak, founded in 1934, is the largest BG/FG music service provider in the United States. On January 14, 2005, ASCAP and Muzak entered into a blanket license for the period January 1, 2005 through December 31, 2009 (the "Muzak Agreement"). In exchange for this license, Muzak agreed to pay ASCAP a flat fee of [REDACTED] per year for each of the five years of the license term, regardless of whether the number of subscribers to the service decreased. As the agreement explains, "Muzak assume[d] all risks associated with the possibility of [a] declining subscriber count[]." The agreement indicates that as of October 31, 2004, Muzak had no more than [REDACTED] locations.[11] A public performance fee of [REDACTED] for [REDACTED] locations yields an effective rate of $41.21 per location—roughly a 16% drop from the $49.50 rate that the BG/FG industry had paid for 5A locations since the final year of the 1994 Form Agreement. The Muzak Agreement also regulated the fee in the event that Muzak increased its subscriber base. The parties agreed that the annual fee would not increase unless the number of Muzak locations increased at a rate exceeding 8% per year. Specifically, for each of the locations in excess of the number of locations that would have equaled 8% of the prior year's total number of locations, Muzak agreed to pay ASCAP an additional annual fee of $41 per location. If over the course of the license term Muzak increased its locations each year at a rate of just 8% per year, by the final year of the license term, the effective per location rate would have dropped as low as $28.05 per location. Under that same scenario, the average fee for the five year period would have been $32.52. As the terms of the Muzak Agreement reflect, Muzak expected that it would continue to expand its business, as it had been doing during the first years of the new century. For its part, ASCAP also anticipated that Muzak's business would continue to grow.[12] When ASCAP entered the *525 Muzak Agreement, it expected that it would become the template for final fees for the BG/FG music services industry, including for several large competitors of Muzak that ASCAP expected would grow significantly in the coming years. Consequently, as of early 2005, ASCAP agreed to a license that it expected to result in a declining per location rate for the five year period ending in 2009. ASCAP and Muzak also addressed both the interim license fees that Muzak owed ASCAP from 1999—that is, the period that followed the expiration of the 1994 Form Agreement—and an outstanding audit dispute between the parties. According to the Muzak Agreement, the interim fees that Muzak had already paid to ASCAP during this period were considered final. With respect to the audit dispute, Muzak agreed to pay ASCAP [REDACTED] to settle a [REDACTED] claim brought by ASCAP for the period beginning December 2002 and ending March 2004. This settlement amount was not allocated to any particular year within the settlement period. If these two settlements are considered part of a single transaction with the Muzak Agreement—as the contracts and circumstances suggest they should— they also reflect that ASCAP and Muzak agreed to an effective per location rate of substantially less than $41.21. But, the Muzak business did not grow. To the contrary, Muzak decreased in size, reaching a low of [REDACTED] locations in 2009. On April 24, 2009, ASCAP agreed to reduce Muzak's annual fee to [REDACTED], in light of Muzak's then-pending bankruptcy. Since the expiration of the Muzak Agreement on December 31, 2009, Muzak has been paying ASCAP an interim flat fee of [REDACTED] per year, subject to retroactive adjustment when the parties reach an agreement as to final fees.[13] 4. 2010 Music Choice Agreement Although it is not nearly as large as Muzak, Music Choice is another leader in the BG/FG music services industry. It provides its services through cable and satellite television systems. In December 2004, Music Choice served [REDACTED] locations and by the end of 2009, it provided music to [REDACTED] establishments. Like Muzak, from 1994 until 1999, Music Choice operated under ASCAP's 1994 Form Agreement. During the Muzak proceedings, Music Choice and ASCAP entered into a letter agreement dated May 15, 2003, which provided that Music Choice would be "bound" by all license terms that ASCAP, Muzak, and DMX Music, Inc. agreed upon or that were determined by the rate court. After the Muzak Agreement was executed in 2005, Music Choice and ASCAP engaged in negotiations over the application of the Muzak Agreement to their relationship. ASCAP initially offered a license at a rate of $46.00 per location, and then a rate of $41.21 per location. Music Choice rejected both proposals. The parties also disputed how the 8% organic growth provision in the Muzak Agreement should be applied to Music Choice.[14] The effective per location rate would have been $35.71 *526 under ASCAP's interpretation of the growth provision and only $28.79 if Music Choice's construction were adopted. Unlike Muzak, at the time Music Choice was negotiating its final license fee with ASCAP it knew how its business had grown. Consequently, there was no uncertainty as to the extent to which Music Choice would increase its number of customer locations during the license period. As the negotiations reached an impasse, Music Choice stopped paying interim fees, asserting that it had overpaid ASCAP in prior years. Thus, from July 2008 until September 2009, it made no payments to ASCAP. On May 21, 2010, ASCAP and Music Choice settled their outstanding fee disputes for the prior decade by adopting a final fee (the "2010 Music Choice Agreement"). As is true with most settlements, this agreement was a compromise. Music Choice paid ASCAP [REDACTED] beyond the interim fees it had already paid "in full and final settlement of all claims either party shall have or may have against the other ... for additional license fees or other amounts owed to ASCAP for the Music Choice Service for all periods through December 31, 2009."[15] The parties in this rate court proceeding contest the effective rate that can be fairly attributed to the settlement of the Music Choice fee dispute covering the ten and a half year period between June 1, 1999 and December 31, 2009. DMX calculates an effective rate of either $29.89 for the most recent five year period, or $34.60 per location for the entire ten year period. While ASCAP calculated an effective rate of $40.36, it did not show that that figure was based on any sound methodology. In sum, in 2010, at a point in time when ASCAP and Music Choice knew precisely how Music Choice's business had expanded in recent years, the parties agreed to settle their fee dispute for a per location rate significantly below $41.21. 5. 2009 PlayNetwork Agreement PlayNetwork is another leader in the BG/FG music services industry. It grew from 12,184 locations in December 2004, to 49,110 in 2009. Its major customers include chains such as Starbucks. On August 14, 1998, PlayNetwork signed a 1994 Form Agreement, granting PlayNetwork a blanket license from its creation in 1996 through May 31, 1999. When this license expired, PlayNetwork agreed to pay ASCAP interim fees on the basis of the rates that applied to it in May 1999 under the 1994 Form Agreement. Like Music Choice, on May 15, 2003, PlayNetwork signed a letter agreement with ASCAP stating that it would be bound by the terms of the agreement that ASCAP reached in its then-ongoing rate dispute with Muzak and DMX Music, Inc. On May 8, 2009, ASCAP and PlayNetwork entered into an agreement (the "2009 PlayNetwork Agreement") to resolve their disputes for the period 1999 through 2009. It provided that interim fees paid between 1999 and 2004 would "be deemed to be final license fees." During this time frame, PlayNetwork served on average, approximately 5,000 locations. ASCAP calculates that this settlement for the five-and-a-half year period is equivalent to a per location rate of $49.43. The 2009 PlayNetwork Agreement also provided that for the period beginning January 1, 2005 and ending December 31, 2009, PlayNetwork would pay a license fee at a rate of $41.21 *527 per location.[16] Thus, the 2009 PlayNetwork Agreement is a recent license agreement with a major BG/FG music service provider, albeit one about half of DMX's size, that adopts a $41.21 rate. 6. 2009 Trusonic Memorandum of Understanding Trusonic is the smallest of the BG/FG music service companies that have more than 10,000 customer locations. While in 2004, it delivered music to only 2,259 locations,[17] by 2009, this number had grown to 11,487. Trusonic has several large chain accounts, such as Toys "R" Us and Crate & Barrel. Trusonic differs from its competitors in the BG/FG service industry in that certain Trusonic customers only receive music from publishers who have entered into direct licensing agreements with Trusonic. On January 29, 2003, Trusonic and ASCAP entered into an interim license agreement under which Trusonic agreed to pay ASCAP the May 1999 rates under the 1994 Form Agreement. In a letter dated November 13, 2006, ASCAP notified Trusonic of an audit finding that Trusonic was almost $1.1 million behind in its fee payments. Trusonic disputed the amount it owed ASCAP. On March 24, 2009, the parties executed a memorandum of understanding ("MOU") to resolve their dispute regarding the 2003 through 2007 period and to provide a framework within which to negotiate a license for the period beginning January 2008 and ending December 2012. The MOU set a final fee for the period from February 1, 2003 through December 31, 2007, at $500,000 over the amount Trusonic had already paid ASCAP. For the 2008 through 2012 period, Trusonic agreed to pay ASCAP at an annual rate of $45 per location, but only for those establishments that play ASCAP music that is not otherwise licensed. In other words, Trusonic will not pay ASCAP anything for locations that only play music within the ASCAP repertory that Trusonic has directly licensed from publishers, or that play no ASCAP music. The MOU further provides that the parties "will negotiate in good faith and execute a license agreement" for this latter period, but they have not yet done so pending the outcome of this rate court proceeding. ASCAP tracks the number of Trusonic premises that play ASCAP music which is not directly licensed, but it does not know the total number of Trusonic locations. Consequently, it is impossible to calculate the actual per location rate that Trusonic is paying ASCAP. Presumably, this rate is less than $41.21, since during the negotiations preceding the adoption of the $45 rate, Trusonic rejected ASCAP's proposal that it pay $41.21 for all locations. D. Competitive Environment in the Background/Foreground Music Services Industry In addition to DMX's established competitors, new forms of competition have emerged in the BG/FG music services industry. "Music consultants" including Activaire, Audiostiles, Gray V, and Music Styling, have replaced traditional BG/FG music service providers by using iPods or other storage and playback devices to create music playlists that are customized for particular applications. Similarly, business owners are increasingly using their own music storage devices to create their *528 own playlists. Even more recently, Internet-based commercial streaming services have begun to enter the market. This increase in competition has affected the fee that DMX can charge its customers as well as its revenue.[18] Both have declined. In the roughly two year period between July 2008 and September 2010, the rates DMX charges its customers have declined by 25%.[19] In the five year period since June 2005, DMX's average per-customer recurring monthly revenue has declined from approximately $57 in June 2005 to $36-a 37% decrease. Recently, DMX successfully bid for a contract with a retailer with 900 locations. It offered its BG/FG music service, including installation of new equipment, at a rate of $24.75 per location per month. E. DMX's Direct License Program In early 2006, DMX began an initiative to secure the right to publicly perform musical compositions directly from music publishers. In so doing, DMX sought to reap the benefits of Muzak I's holding that ASCAP is required to offer music users blanket licenses with carve-outs. The fees that DMX pays to the three PROs—ASCAP, BMI, and SESAC—for blanket through-to-the-listener licenses constitute one of DMX's largest costs of sale and DMX is hoping to reduce those costs as it confronts declining per location revenues in an increasingly competitive market.[20] Because DMX's programs are designed around music genres or themes, it is largely liberated from the need to provide its customers with any particular song. As a result, DMX is able to construct programs that rely heavily on music covered by its direct licenses without adversely affecting the quality of the programming. As of the time of trial, DMX had executed more than 850 direct licenses that accounted for roughly 30% of its programming. In January 2006, DMX engaged MRI, a company specializing in high-volume music license administration, to assist in the design and implementation of the direct licensing campaign. MRI was created in 1989 to help radio and television broadcasters take advantage of the per-program license available under the PRO consent decrees. To that end, MRI developed systems for tracking the broadcasting of television programs, analyzing their music content for PRO-affiliated compositions, and preparing music usage reports for the PROs on behalf of per-program licensees. More recently, MRI has capitalized on its expertise in the licensing and administration of music rights by marketing its services as an experienced negotiator of direct licensing agreements. Because it is important to MRI that it maintain good relationships with music publishers, MRI *529 developed a DMX proposal that it believed would be credible and well received by the music publishing community. Initially, MRI conducted a "feasibility study," which analyzed one month of DMX data to identify those publishers whose works DMX played most frequently. DMX and MRI then developed a "Music Composition Catalog License" ("MCCL")—a generic direct license agreement that was intended to be readily acceptable to a wide range of publishers, without the need for extensive, individual negotiations. MRI tried to include provisions in the MCCL that would be familiar to music publishers in order to encourage their participation and reduce the need for lengthy negotiations. A high degree of uniformity across all direct license agreements was necessary to ensure that DMX would be able to efficiently administer its royalty obligations under each of the individual licenses. After identifying the fifty or so publishers on whose compositions DMX relied most heavily, MRI began to negotiate with them to get as many MCCLs executed as possible. Rather than targeting those publishers who would be most likely to execute an agreement, MRI first approached more sophisticated publishers, so that the form license could be modified in response to their feedback. In the course of these negotiations, MRI explained, in essence, that DMX was trying to maximize its reliance on direct licensing and minimize its reliance on the ASCAP blanket license. As a result, those publishers who gave DMX a direct license could expect that DMX would play more of their music, and those who did not could expect that DMX would play less. MRI used this simple appeal to publishers' financial interest to get as many MCCLs executed as possible.[21] In the course of negotiating the MCCL with the first fifty publishers and then with the next fifty or so who accounted for the largest share of DMX music, DMX occasionally agreed to alter individual provisions in the MCCL, either for a single publisher or as a change to the form MCCL. Nonetheless, the MCCL terms have remained fairly stable and the parties have identified a typical MCCL. After laying out the relevant terms of the standard MCCL, the remainder of this section describes the current state of DMX's direct license initiative and the steps that DMX has taken to secure the participation of "major" publishers, most notably Sony. 1. Key Terms of the MCCL a. Scope of Rights Provision The MCCL provides DMX with a broader scope of rights than it would otherwise receive under an ASCAP public performance license agreement. Not only does an MCCL authorize DMX to publicly perform the musical works in a publisher's catalog, but it also grants DMX a non-exclusive license to reproduce, distribute, and edit[22] such works. Additionally, the MCCL extends to DMX's delivery of music to all kinds of business establishments, including businesses that charge admission.[23] Both ASCAP and BMI typically exclude from *530 their form blanket licenses the right to publicly perform works in venues that charge admission. b. Twenty-Five Dollar Royalty Rate Under the MCCL, publishers receive a pro rata share of a royalty pool. This pool is calculated by multiplying the number of DMX locations by a fee of $6.25 per quarter, or $25 per year. The chairman of MRI and a DMX executive independently arrived at the $25 rate as an appropriate rate, and DMX adopted it. Gertz, a thirty-five year veteran of the music industry with an extensive background in the negotiation and administration of copyright licenses, reached the $25 rate based on his experience negotiating a similar direct-license arrangement with Sony and EMI Music Publishing ("EMI") on behalf of Muzak. As a result of these talks, Sony and EMI had consented to a $28 per location royalty pool. These agreements were never implemented, however, because Muzak ultimately entered blanket licenses with ASCAP and BMI in 2005 that did not contain carve-out fee mechanisms. See discussion of Muzak Agreement, supra. Knittel, whose long career at ASCAP before he joined DMX's predecessor included negotiating licenses with large industries, began his calculation with a benchmark of 2.25% of revenue. This was the rate BMI was then charging the BG/FG music services industry for off-premise delivery of music. The MCCL further provides that the royalty pool covers both the publisher's and the composer's rights in DMX's use of a musical composition. Under the MCCL, only the publisher will be paid by DMX and the publisher has the duty to distribute to the composer her share of the royalty.[24] c. Royalty Formula A publisher's distribution from the royalty pool is calculated by multiplying the total pool by a fraction which isolates the percentage of the pool to which a publisher is entitled in a given period. The numerator of this fraction is the number of times that a particular publisher's works were performed on DMX's off-premise service. The denominator is the total number of performances on the off-premise service. The MCCL illustrates how this formula functions: [I]f one hundred thousand (100,000) Locations exist for the duration of a quarter, and Publisher's pro rata share for that quarter equals one percent (1%), the Publisher will receive royalties for that quarter equal to Six Thousand Two Hundred Fifty Dollars ($6,250.00) ($6.25 × 100,000 × 1%). Finally, if a publisher owns less than the entirety of a given composition, it will only be entitled to royalties based on the percentage of the work it controls. Thus, the numerator of the above-described fraction is adjusted to reflect the publisher's actual ownership interest in a given work. d. Performance Proxy All locations that DMX serves are taken into account when calculating the total royalty pool, however, only those locations served via off-premise DBS transmissions are considered when determining a publisher's pro rata share of the total pool. DMX chose to use the frequency that a work is played on the off-premise channels as a proxy for the number of times a composition is played at on-premise locations because, while the two forms of service are roughly equivalent in terms of programming content, DMX has better *531 data regarding the works played at off-premise locations. e. No Double Payment Provision A "No Double Payment" term in the MCCL provides that if a PRO (or any music collection society administering mechanical rights) makes a claim against DMX for the performance of a work that is licensed under a MCCL, DMX will not be obligated to pay the publisher if DMX "has paid" the PRO for the right to use the composition. It reads as follows: No Double Payment: Publisher acknowledges that, during the Term, Licensee may be a licensee of ASCAP, BMI, and/or SESAC (each, a "Music Collection Society"). From time to time, a Music Collection Society may make a claim against Licensee for payment of royalties for the public performance of Compositions otherwise licensed hereunder. Notwithstanding anything to the contrary in this Agreement, Licensee will have no obligation to pay Publisher with respect to the public performance rights for any Composition during any period of the Term where Licensee has paid a Music Collection Society for such rights to such Composition in such period. (Emphasis supplied.) The text of the No Double Payment provision emerged from MRI's experience implementing per-program licenses in the television industry. In that context MRI reports performances to ASCAP and BMI on a performance-by-performance basis. At times, ASCAP and BMI will dispute that a direct license covers a particular performance, and there is a process for reconciling those disagreements. f. Most Favored Nations Clause The "Most Favored Nations" clause ("MFN") of the MCCL assures publishers who have entered into direct license agreements with DMX that their royalty will be calculated using the same formula that DMX uses to calculate the royalties for all other publishers, i.e. that no publisher will be entitled to a greater percentage of the pool than the ratio derived by dividing the number of the directly licensed publisher's plays by the total number of DMX performances. It reads as follows: Most Favored Nations: Licensee represents and warrants that the Royalty Formula will be the same as the Royalty Formula for calculating royalties payable to any third party music publisher or administrator (a "Competing Publisher"), under any catalog license between Licensee and any such Competing Publisher which is entered into during the Term, for the same rights, Territory, and Term as provided herein. For the avoidance of doubt, "Competing Publisher," as used herein, will not include a Music Collection Society. ASCAP contends that the reference to "royalty formula" in the MFN provision also requires DMX to offer all licensors a comparable "royalty pool," to wit, a share of the same pool of money created by the $25 per location rate, or that at the very least it would have been understood by publishers to include such a commitment. The text of the MCCL is less than artful with respect to this point, but there is some evidence that the MCCL's MFN provision does not encompass the royalty pool. In response to the demands of one music publisher—Cherry Lane—DMX modified the MFN in one MCCL to provide that all publishers will be treated the same for the purposes of both the royalty formula and the per location rate, i.e., $25 per location. It is unnecessary to determine whether the MFN provision creates a contractual commitment by DMX to give every publisher a proportionate share of the $25 per location rate, however, since DMX has incorporated *532 the $25 per location royalty rate in all of its direct license agreements to date. The scope of the MFN has particular relevance to the treatment of the advance that DMX paid to Sony, which is described below. There is little indication in the record that publishers concluded that the MFN would encompass advances paid to music publishers. Several publishers testified that they did not interpret the MFN as covering advances. 2. The MCCL Program from 2006 to 2010 In May 2006, Leiber Stoller Songs—a publisher whose song list includes, "Stand By Me," "Spanish Harlem," "There Goes My Baby," "Charlie Brown," and "Jailhouse Rock"—became the first publisher to enter into a direct licensing agreement with DMX. The three-year license term began on October 1, 2006. The Leiber-Stoller catalog has since been purchased by Sony and remains directly licensed, but pursuant to DMX's agreement with Sony (discussed at length below). Since signing Lieber-Stoller, DMX has entered into hundreds of MCCL agreements.[25] As of September 10, 2010, DMX had secured approximately 850 direct licenses with publishers representing a total of more than 7,000 catalogs. Many of these direct license agreements are with publishers who represent a single catalog or the works of a single songwriter. Others are signed by administrators acting on behalf of multiple music publishers. These licenses give DMX access to a diverse array of musical works that vary in terms of genre, time period, and even language.[26] Additionally, DMX's direct licenses extend to works that are currently being performed by popular contemporary artists, including Beyoncé, Alicia Keys, and Carrie Underwood. At the 2010 Grammy Awards, for instance, the award for "Song of the Year" went to an artist performing a composition that is directly licensed by DMX: "Single Ladies (Put a Ring On It)." 3. Renewals of MCCLs Not only has DMX been able to enter into new MCCL agreements, but it has also had significant success in renewing agreements with those publishers whose initial licenses have expired. The typical term for an MCCL is three years. Seventy-eight out of the ninety-two publishers whose licenses have come up for renewal have renewed either by express authorization or pursuant to an automatic renewal provision in the initial agreement.[27] *533 4. Refusals to Renew MCCLs Of the fourteen publishers that did not renew MCCLs, at least three controlled significant catalogs: The Walt Disney Company, Abkco Music, Inc., and Cherry Lane. BMI contacted all three publishers. Indeed, BMI told Cherry Lane that if it renewed its direct license agreement with DMX, BMI would force Cherry Lane to repay BMI the payments BMI had made to it during the direct license period. These three publishers accounted for just over 2% of the music performed by DMX and its licensees. A fourth publisher declined to renew after agreeing to serve as a BMI witness in BMI's litigation with DMX.[28] Of the remaining ten publishers, six were small catalogs that were sold to a new administrator that had not signed an MCCL agreement. The final four were catalogs represented by an administrator that renewed its MCCL agreement with DMX, but only for its other catalogs. 5. DMX's Outreach to "Major" Publishers: The Sony MCCL There are four major publishers in the music publishing industry that control a considerable share of the total market: Sony, Warner/Chappell, EMI, and Universal. From the early stages of the development of its direct license initiative, DMX recognized that in order for the program to be successful, DMX would need to secure the participation of at least one major publisher. Indeed, in response to DMX's solicitations to enter into MCCL agreements, numerous independent publishers indicated that they would be willing to execute the MCCL if DMX signed some number of the majors.[29] Given the strong affiliation between the PROs and the major publishers,[30] DMX believed that in order to conclude an MCCL agreement with a major it would be necessary to offer an attractive royalty advance. On November 14, 2007, Sony became the first and, to date, the only major publisher to enter an MCCL agreement with DMX.[31] Sony promptly advised ASCAP that it had executed the MCCL.[32] The Sony MCCL *534 grants DMX the through-to-the-listener public performance rights and mechanical rights to the entire Sony repertory, except for the Neil Diamond catalog. The fee structure in the Sony MCCL is identical to the rate provisions in all of DMX's direct license agreements: Sony is entitled to its pro rata share of the annual royalty pool, which is calculated by multiplying $25 by the number of DMX locations. The Sony license is distinguishable from all of DMX's other direct license agreements, however, in that the Sony MCCL provides for a $2.4 million "non-returnable, non-refundable advance" that is "recoupable from any and all royalties payable to [Sony]." Additionally, in a letter agreement signed the same day as the MCCL, DMX agreed to pay Sony $300,000 to cover the "additional administrative and overhead costs" that Sony would "incur in connection with the Agreement." This supplemental fee was also "recoupable from any and all royalties owed to Sony by DMX under the Agreement." Thus, DMX gave Sony a $2.7 million recoupable, but non-refundable advance. Under the license, either party can publicly disclose the existence of the agreement, but most of its terms—including the advance—are confidential. DMX employed a four-step formula to calculate the Sony advance. First, DMX identified the amount of money that Sony had received from ASCAP and BMI on behalf of DMX in the prior year-approximately $300,000. This amount was then doubled, because unlike the payment that Sony received from the PROs, the DMX royalty fee covers both the composer's and the publisher's share. Next, this figure—$600,000—was increased by fifty percent, resulting in an annual rate of $900,000. This 50% increase reflects the premium DMX was willing to pay to get a major publisher to sign a direct license. Finally, to determine the advance for the entire license term, the annual rate was multiplied by three-the number of years in the agreement-for a total advance of $2.7 million. Although DMX intended to recoup as much of the advance as possible, by expanding its locations and increasing its reliance on the Sony catalog, DMX could not have realistically expected to recoup the full advance. DMX suspected that Sony had quoted DMX an inaccurate figure for the amount of royalty payments it had received from BMI due to music performed by DMX and its customers. This intuition was confirmed in October 2008, when discovery taken in the rate court proceeding between DMX and BMI revealed that the number quoted to DMX included BMI payments not just for music performed by DMX but for all of the music performed by the entire BG/FG music services industry.[33] Only 27% of the royalties on the "DMX" line were actually attributable to DMX performances of the Sony repertory. On April 9, 2009, DMX and Sony signed an amendment to the Sony MCCL (the "2009 Amendment") that extended the period in which DMX could recoup its advance for an additional twenty-one months. While DMX contends that Sony granted it this extension solely to compensate DMX for its miscalculation of the advance, there is evidence suggesting that Sony had been considering giving DMX an extension prior to learning about the confusion surrounding the BMI royalty statement.[34] *535 Due to the twenty-one month extension that Sony granted DMX in the 2009 Amendment, the period in which DMX can recoup its advance from Sony will not end until September 30, 2012. The parties dispute the extent to which it will be possible for DMX to recoup this initial investment as a credit against royalty payments. Because DMX has control over its programming, it has already doubled the percentage of its catalog that is derived from Sony music, increasing it from roughly 6% to 12%. In addition, to the extent DMX continues to add to its revenue base by securing more locations, it also increases its recoupment of the advance. Nonetheless, there is little to no likelihood that the Sony advance will be entirely recouped by September 30, 2012. 6. DMX's Negotiations with Other Major Music Publishers DMX has reached out to the three other major music publishers, but for several reasons, has not concluded MCCL agreements with any of them. The PROs have tried their best to prevent the major music publishers from entering direct licenses with DMX. ASCAP CEO John LoFrumento actively tried to discourage the CEOs of Universal and Sony from entering into MCCL agreements. BMI offered Universal a substantial guaranteed advance payment if it refrained from signing an MCCL. In addition to these efforts by PROs, at least two of the major publishers were concerned that the $25 per location royalty pool was too low. F. DMX's Application for an ASCAP Blanket License DMX requested a through-to-the-listener blanket license from ASCAP on June 3, 2005. On November 21, 2005, ASCAP provided DMX with its first quote—a per location annual rate of $48 to $52, depending on the number of DMX subscribers. DMX rejected this proposal. On July 25, 2006, ASCAP initiated this rate court proceeding by applying for the determination of a reasonable fee. ASCAP offered DMX a revised proposal on October 3, 2006, based on the Muzak Agreement. Under the second proposal, the interim fees that DMX had already paid ASCAP for the period between June 2005 and December 2006 would be deemed final, and DMX would be charged at a $41.21 per location rate for the period beginning September 2006 and ending December 2011. Again, DMX rejected ASCAP's proposal. On June 13, 2007, the parties settled on a new, lower interim rate, retroactive to January 2007. They could not, however, reach agreement as to a final rate. At a court conference held on August 20, 2009, ASCAP acknowledged that under Judge Conner's Muzak I decision, it was obligated to "account for [DMX's] direct licensing in formulating [its fee] proposal." ASCAP's final fee proposal, which is described at length below, was delivered to DMX on April 5, 2010. Contrary to its representation in August 2009, ASCAP's fee proposal did not account for DMX's direct licensing initiative in any meaningful way. G. DMX'S Application for a BMI Blanket License At the same time that DMX applied to ASCAP for a blanket license in 2005, it also requested a blanket license with carve-outs from BMI. Unlike ASCAP, BMI did not object to the structure of the requested license: the parties agreed, in principle, to an adjustable fee blanket license *536 under which DMX would receive a proportional credit off the blanket license fee to reflect performances of BMI music that were directly licensed. They also agreed that the fee would be constructed from three components: a Floor Fee, a Blanket Fee, and a Direct License Ratio. BMI and DMX could not agree, however, on the benchmarks upon which the Floor Fee and the Blanket Fee should be based. They also disputed which performances should be considered when calculating the Direct License Ratio. On January 10, 2008, BMI filed a petition with the BMI rate court. Broadcast Music, Inc. v. DMX, Inc., 726 F.Supp.2d 355, 356-57 (S.D.N.Y.2010) ("BMI/DMX"). On July 26, 2010, the BMI court entered a final rate for the blanket license, subject to adjustment based on the proportion of DMX's BMI performances that are directly licensed. Id. at 364-65. Based on the record created at a February 2010 trial, the Honorable Louis L. Stanton chose the $25 royalty pool rate in the MCCLs as the relevant benchmark for determining the value of the public performance of music.[35]Id. at 360-62. Judge Stanton also found that the Sony advance was a "cost of entry" into the market rather than a royalty payment, and therefore the $25 royalty pool in the MCCL did not need to be adjusted to account for the portion of the Sony advance that DMX will be unable to recoup. Id. at 361-62. BMI raised a number of issues during its trial with DMX, however, that have not been raised by ASCAP in the instant proceeding. These include various arguments relating to the Floor Fee and a request for an increase in the Blanket Fee due to the "option value" that the adjustable fee blanket license provides DMX. Id. at 361-64. Based on his resolution of all of the issues presented, Judge Stanton approved a per location fee of $18.91. Id. at 364-65. The structure of DMX's proposal in this proceeding is similar to the one to which BMI consented, and DMX's proposed benchmark for the valuation of the right to publicly perform music is the same as that adopted by Judge Stanton. At the DMX/ASCAP trial, ASCAP's Director of Licensing took the position that to the extent that ASCAP and BMI have comparable market shares, adopting the BMI rate is a "proper way to determine the [ASCAP] fee[]." It is not disputed that, in fact, the ASCAP and BMI market shares of DMX performances are roughly equal. CONCLUSIONS OF LAW After a brief description of the kinds of licenses that ASCAP must offer music users under AFJ2 and the standard for determination of a reasonable fee, this Opinion will examine the two fee proposals submitted by ASCAP and conclude with a review of the DMX proposal. The Opinion finds that ASCAP is required to issue DMX a blanket license with carve-outs for DMX's direct licensing program. After finding that neither of ASCAP's fee proposals is reasonable, the Opinion adopts DMX's proposal as representing a reasonable fee for the right to publicly perform music licensed by ASCAP. A. The Governing Law Under AFJ2, ASCAP is required to grant music users who make a written request a "non-exclusive license to perform all of the works in the ASCAP repertory." AFJ2, § VI. AFJ2 describes two licenses of interest to these proceedings. The first is a blanket license, which is defined as a "non-exclusive license that authorizes a *537 music user to perform ASCAP music, the fee for which does not vary depending on the extent to which the music user in fact performs ASCAP music." Id. § II(E). The other license—the per-segment license—is a non-exclusive license to perform any or all works in the ASCAP repertory in "all segments of the music user's activities in a single industry, the fee for which varies depending upon which segments contain ASCAP music not otherwise licensed for public performance." Id. § II(K). ASCAP must "advise the music user in writing of the fee that it deems reasonable for the license requested. ..." Id. § IX(A) (emphasis supplied). If the parties are unable to reach agreement, either of them may apply to the Court for a determination of a reasonable fee. Id. AFJ2 "gives ASCAP the initiative in proposing the entire formula" for calculating a reasonable fee. United States v. ASCAP (In re Application of Metromedia), 341 F.2d 1003, 1009 (2d Cir.1965). But ASCAP is only required to respond to requests for the kinds of licenses "which some other portion" of the consent decree requires ASCAP to grant. United States v. ASCAP (In re Application Shenandoah Valley Broadcasting, Inc.), 331 F.2d 117, 122 (2d Cir.1964) ("Shenandoah Valley").[36] In AEI, 275 F.3d 168, the Second Circuit considered whether the rate-setting provision of the BMI Consent Decree permits the rate court to enter a "blanket license subject to carve outs, i.e., a license that would enable a licensee to reduce its fee obligation to BMI to the extent it had licensed works represented by BMI directly with BMI's composer and music publisher affiliates." Id. at 173 (citation omitted). The AEI court concluded that under the BMI Consent Decree, BMI was obligated to offer an applicant "a blanket license with a fee structure that reflects such alternative licensing" and that its "[f]ailure to do so will empower the district court to set a reasonable fee." Id. at 177. This holding was based on two findings of particular relevance to the DMX application. First, the court found that blanket licenses are a form of license that BMI must make available to music users under the BMI Consent Decree. Id. at 176. Second, the court held that a blanket license subject to carve-outs "differ[s] from the traditional blanket license only in its fee structure," and is therefore still a "blanket license" within the meaning of the consent decree. Id. at 177. The provision of the BMI Consent Decree requiring the PRO to respond to a request for a license is almost identical to the corresponding provision in AFJ and AFJ2.[37]See Muzak I, 309 F.Supp.2d at 577. In Muzak I, this court followed the Second Circuit's holding in AEI and found that under § IX(A) of the AFJ2, the existence of direct licensing relationships may and will be considered in a rate court proceeding under AFJ2 section IX in determining whether ASCAP has met its burden of proving the reasonableness of the blanket licensing fee it seeks or, in the event that ASCAP fails to meet that burden, in the Court's *538 calculation of a reasonable fee based on all the evidence. Id. at 581 (emphasis supplied). In any fee-setting proceeding, ASCAP bears the burden of showing that the fee it seeks is reasonable, absent exceptions not relevant here. AFJ2 § IX(B). If ASCAP does not establish the reasonableness of its proposed fee, "the Court shall determine a reasonable fee based on all the evidence." AFJ2 § IX(D). Although AFJ2 does not define what is meant by a "reasonable fee," governing precedent dictates that in determining the reasonableness of a licensing fee, a court "attempts to make a determination of the fair market value—the price that a willing buyer and a willing seller would agree to in an arm's length transaction." United States v. BMI (In re Application of Music Choice), 316 F.3d 189, 194 (2d Cir.2003) ("Music Choice II") (quoting ASCAP v. Showtime/The Movie Channel, 912 F.2d 563, 569 (2d Cir.1990) ("Showtime")). Fair market value is a "hypothetical" matter. Showtime, 912 F.2d at 569. "[S]ince there is no competitive market in music rights," the parties and courts ordinarily "lack any economic data that may be readily translated into a measure of competitive pricing for the rights in question." Id. at 577 (district court opinion). Thus, when setting a reasonable fee, a court should consider "the context in which [the rate court] was created." Id. at 570. Assessing a reasonable rate necessarily requires a court to "tak[e] into account the fact that ASCAP, as a monopolist, exercises disproportionate power over the market for music rights." United States v. ASCAP (In re Application of RealNetworks, Inc. and Yahoo!), 627 F.3d 64, 76 (2d Cir.2010) (citation omitted) ("RealNetworks/Yahoo!"). Therefore, the rates established by its pattern of licensing must be assessed carefully. After all, [t]he opportunity of users of music rights to resort to the rate court whenever they apprehend that ASCAP's market power may subject them to unreasonably high fees would have little meaning if that court were obliged to set a `reasonable' fee solely or even primarily on the basis of the fees ASCAP had successfully obtained from other users. Showtime, 912 F.2d at 570. In sum, "[f]undamental to the concept of `reasonableness' is a determination of what an applicant would pay in a competitive market." RealNetworks/Yahoo!, 627 F.3d at 76. A determination of the fair market value "is often facilitated by the use of a benchmark—that is, reasoning by analogy to an agreement reached after arms' length negotiation between similarly situated parties." Music Choice II, 316 F.3d at 194. When choosing a benchmark and deciding how that benchmark should be adjusted for the application it is considering, a rate court must determine the degree of comparability of the negotiating parties to the parties contending in the rate proceeding, the comparability of the rights in question, and the similarity of the economic circumstances affecting the earlier negotiators and the current litigants, as well as the degree to which the assertedly analogous market under examination reflects an adequate degree of competition to justify reliance on agreements that it spawned. United States v. BMI (In re Application of Music Choice), 426 F.3d 91, 95 (2d Cir. 2005) ("Music Choice IV") (citation omitted). Finally, AFJ2 enjoins ASCAP from engaging in certain prohibited conduct. Among other things, ASCAP is restrained from "[l]imiting, restricting, or interfering with the right of any member to issue, *539 directly or through an agent other than a performance rights organization, non-exclusive licenses to music users for rights of public performance." AFJ2 § IV(B). B. ASCAP's Rate Proposals ASCAP has presented two proposals for a DMX blanket license. Its first and preferred proposal is for a blanket license with no carve-out for DMX's direct licensing program. In the event that the Court finds this proposal to be unreasonable, it has tendered as an alternative a blanket license with a static credit based largely on the payments DMX made to its direct licensors in 2009. In presenting these two proposals to DMX, ASCAP has refused to quote a reasonable fee for a blanket license that would accommodate DMX's direct licensing program. This refusal stems from ASCAP's contention that a blanket license with a carve-out for direct licensing is not a reasonable fee structure, and that if DMX wanted anything other than the traditional blanket license, it should have requested a per-segment license. 1. ASCAP's First Proposal: A Flat Fee Blanket License There are two parts to ASCAP's first proposal. First, for the period beginning June 5, 2005 and ending December 31, 2009, DMX would be charged a flat fee of $15,677,777 for its performances of all works in the ASCAP repertory by means of its BG/FG music service. Second, for the period beginning January 1, 2010 and ending December 31, 2012, DMX would pay $49 per location, which is equivalent to the effective rate of the DMX license for 2009, based on calculations derived from the first prong of this proposal. To take account of inflation, this rate would be adjusted in 2010 and 2011 to reflect changes in the Consumer Price Index ("CPI"). ASCAP contends this first proposal is "proportional" to the fees that Muzak agreed to pay ASCAP for a blanket license for its BG/FG music service in 2005, the same year that DMX requested a license. To arrive at the $15,677,777 figure, ASCAP begins with the [REDACTED] annual fee that Muzak agreed to pay ASCAP in 2005. ASCAP divides the [REDACTED] total fee by the number of locations that Muzak serviced in 2005—[REDACTED]— arriving at a per location rate of $41.21. ASCAP then multiplies the per location rate of $41.21 by the number of DMX customer locations in 2005-83,000—to arrive at an annual rate of $3,420,606 ($41.21 × 83,000). Since DMX did not request a license until June of 2005, the annual rate for the first year of the term is adjusted so that DMX is only charged for seven months—$1,995,353 ($3,420,606 × 7/12). The total license fee of $15,677,777 is simply the sum of $1,995,353 and $3,420,606 for each of the subsequent four years of the term ($1,995,353 + ($3,420,606 × 4)). To determine the effective rate for 2009, the rate on which the second part of the license is based, the annual flat fee of $3,420,606 is divided by the number of DMX customer locations in 2009—fewer than 69,000. The result is an effective rate in 2009 of $49 per location. ASCAP's preferred proposal is extraordinarily aggressive. It gives DMX no credit whatsoever for DMX's direct licensing program. It also sets a rate far above any yet paid by a licensee. ASCAP's first proposal can be swiftly rejected. Without engaging with the details of the proposal, ASCAP's flat fee blanket license proposal is not reasonable for at least the following two reasons. Pursuant to AFJ2, as interpreted by AEI, 275 F.3d at 173, 177, and Muzak I, 309 F.Supp.2d at 581, ASCAP was required to *540 respond to DMX's request for a blanket license with a carve-out for its direct licensing program by proposing a reasonable rate for that type of license. But, even if the guidance given by these two cases did not impose upon ASCAP the obligation to respond more directly to DMX's request, the facts on the ground required it. As the record established at this trial demonstrates, DMX has a well-established direct licensing program and any reasonable licensing fee would have to take that program into account. ASCAP makes essentially three arguments to support its flat fee proposal. First, ASCAP argues that neither AEI nor Muzak I require it to quote a fee for a blanket license with a carve-out. According to ASCAP, those cases do no more than observe that AFJ2 does not "foreclose the possibility of a blanket license with a discount fee structure." Second, it argues that no willing seller would ever provide a blanket license with a carve-out for a direct licensing program, and therefore, it is not required by AFJ2 or precedent to provide one. Finally, it contends that DMX has an option: if it is unwilling to accept a flat fee blanket license, DMX can apply for a per-segment license. None of these arguments confronts the reality that DMX is entitled to a blanket license when it requests one, and that that blanket fee must accommodate the fact that DMX has a well-developed direct licensing program. ASCAP's quibble about whether AEI and Muzak I either require or only permit a blanket license with a carve-out for a direct licensing program need not detain us long. As explained by the Second Circuit Court of Appeals in AEI, "Applicants' request for a blanket license subject to `carve-outs' constitutes a request not for a new type of license, but for a blanket license with a different fee basis, over which the district court has rate-setting authority and which BMI must offer." AEI, 275 F.3d at 171 (emphasis supplied). Judge Conner held in Muzak I that "the existence of ... direct licensing relationships may and will be considered by this Court in a rate court proceeding under AFJ2 section IX in determining whether ASCAP has met its burden of proving the reasonableness of the blanket licensing fee...." Muzak I, 309 F.Supp.2d at 568 (emphasis supplied). In any event, ASCAP does not dispute that AEI and Muzak I both recognize that a blanket license with a carve-out for a direct licensing program is entirely compatible with AFJ2 and that a court may require ASCAP to issue one. Recognizing that this Court has the authority to grant such a license, it is difficult to understand why ASCAP did not try to shape the contours of such a license by constructing—at least as one of its alternative proposals—a reasonable fee for one. In explaining its refusal to formulate a blanket license fee with a carve-out for direct licensing, ASCAP refers repeatedly to a passage in a brief submitted by the DOJ to Judge Conner in Muzak I. In that passage, the DOJ opined that "[w]hile the AJF2 does not require ASCAP automatically to grant the license request here, it does provide the Court with authority to order such a license." Memorandum of the United States on Decree Construction Issues at 6, Muzak I, 309 F.Supp.2d 566 (emphasis supplied). The license that had been requested there was a license that included a discount for music catalogs that had been directly licensed. The brief concluded that applicants for an ASCAP license "may request, and this Court may order, a carve-out blanket license even under ASCAP's narrow interpretation of the AFJ2." Id. at 19. *541 When DOJ submitted that brief, the parties were wrangling over whether the per-segment license authorized by AFJ2 encompassed the requests by Muzak and DMX Music, Inc. for a license that included a discount for their direct licensing programs. DOJ agreed with ASCAP that a per-segment license was not the appropriate vehicle for the discount. But given its endorsement of a blanket license with a carve-out for direct licenses, the DOJ brief provides little support for ASCAP's position in this trial. It is noteworthy that BMI did not dispute in the rate court proceeding earlier this year that DMX was entitled to a blanket license with a carve-out; the principal dispute between BMI and DMX was over the appropriate fee for such a license. BMI/DMX, 726 F.Supp.2d at 355-56. Indeed, in its very first conference with this Court on August 20, 2009, ASCAP acknowledged that it was "required" to account for the DMX direct licensing program in the fee it would proffer to DMX for a blanket license.[38] Despite that acknowledgement it ultimately refused to provide DMX with any reasonable proposal for such a license. Next, relying principally on the testimony of its expert economist, ASCAP contends that ASCAP cannot be required to issue a blanket license with a carve-out because no willing seller would ever offer such a license. As a result, it argues, such a license is necessarily an unreasonable one and ASCAP is never required under the terms of AFJ2 to issue an unreasonable license. AJF2 is an antitrust consent decree providing a mechanism for the setting of reasonable license fees in a unique market in which ASCAP indisputably exercises market power. While ASCAP may be unwilling to offer a blanket license with a carve-out for a direct licensing program, the terms of AFJ2, the decisions interpreting and applying AFJ2, and the record evidence from this trial each indicate that such a license is appropriate and justified here. Indeed, DMX has shown that such a license will add competition to the marketplace. Finally, as for ASCAP's preference that DMX apply for a per-segment license, that preference cannot override an applicant's option to apply for either type of license. When an applicant applies for a blanket license, AFJ2 requires ASCAP to respond to that request with a proposed reasonable fee for a blanket license. AFJ2 § IX. 2. ASCAP's Alternative Proposal: A Blanket License with a Static Carve-Out ASCAP's alternative proposal builds on its first proposal. After the calculation of the flat fee, that fee is reduced by a credit for payments that DMX has made to ASCAP members for performances of their works licensed through MCCLs. This benefit, however, is offset by an additional charge which ASCAP represents is equivalent to its incremental costs for administering the carve-out license scheme. This basic structure is applied both retrospectively and prospectively. For the period beginning June 5, 2005 and ending December 31, 2009, the blanket rate is the same as the flat fee discussed above: $15,677,777, or $3,420,606 per year. *542 The total royalty payment is reduced by a "dollar-for-dollar" credit for payments DMX made directly to music publishers to license the public performance rights to ASCAP music. To calculate that credit, the amount DMX paid for directly licensed music is reduced by 10% to account for the portion of the fee that is intended to compensate publishers for mechanical rights,[39] and by an additional 50% to identify those royalties that were paid to publishers for compositions that were not in the ASCAP repertory. Finally, ASCAP proposes a charge of $25,000 per year for the additional administrative expense that it will incur in implementing the carve-out license. Since the DMX direct license program did not begin until 2006, DMX would only pay administrative fees for 2007 through 2009, or $75,000. The final fee for this period is calculated by subtracting the discount from the blanket rate and adding the administrative fees: Fee=Blanket Rate—Discount+Administrative Fee ASCAP concludes that DMX should pay $15,410,096 for the June 5, 2005 through December 31, 2009 period. With respect to the period beginning January 1, 2010 and ending December 31, 2012, the blanket rate would be the same as the fee in the first proposal: an annual rate of $49 per location for 2010 and then CPI-adjusted rates in 2011 and 2012. DMX would then be given a $230,000 "carve-out" each year (CPI-adjusted for 2011 and 2012), a figure derived from the amount of royalties that DMX paid to its direct licensors in 2009. Thus, the prospective component of the second proposal does not include a mechanism for taking into account fluctuations in DMX's direct licensing, although DMX's royalty payments will likely change during this period. Finally, the administrative charge would remain constant at $25,000 per year, or $75,000 for the years 2010 through 2012. a. ASCAP Has Not Shown That Its Alternative Proposal is Reasonable For at least four reasons, ASCAP has not carried its burden of showing that its alternative fee proposal is reasonable. First, it has not shown that the per location rate of either $41.21 for the period before 2010 or of $49 for 2010 is an appropriate base from which to construct the blanket rate. Second, ASCAP's proposed flat fee for the years 2006 through 2009 does not account for the fluctuations in the number of locations DMX served in that period. Third, the static discount through which ASCAP proposes to give DMX a credit for its direct licensing program is unreasonable. Finally, ASCAP has failed to provide sufficient evidence to support its imposition of a $25,000 per year charge for administrative costs associated with implementing the credit. The net effect of ASCAP's alternative proposal is that DMX is taxed for having engaged in a direct licensing program: its license fee is higher than it would have been if it had engaged in no such program. It will bear the full cost of its direct licenses and pay an annual administrative fee of $25,000. Simply put, this proposal is designed to convince DMX to discontinue the program: DMX is in a better position if it cancels direct licenses and it is increasingly disadvantaged with every new license it negotiates. As ASCAP's in-house economist candidly admitted at trial, ASCAP's proposals treat DMX "inequitably" since *543 ASCAP believes that DMX should have applied for a per-segment license. i. The $41.21 Per Location Rate The $41.21 per location rate that ASCAP has used to construct its blanket rate rests essentially on a single license: the Muzak Agreement. There are several reasons why it is unfair to calculate a DMX per location rate from that agreement. As described in more detail above, Muzak agreed in 2005 to pay ASCAP flat annual fees of [REDACTED] for each of five years between 2005 and 2009. That agreement, however, included an organic growth provision which gave Muzak the opportunity to achieve a substantially reduced per location rate if its number of locations grew. At the time they entered into the Muzak Agreement, the parties also settled their fee and auditing disputes for the period from 1999 to 2004. Because the parties' agreements resolved issues relating to an eleven year period, and included an organic growth provision, it is dangerous to rely on a per location fee that is calculated from the simple division of the annual [REDACTED] flat fee by the number of locations Muzak serviced in 2005. Moreover, the Muzak Agreement clearly contemplated that Muzak would be expanding its business: it provided a mechanism for increasing its fee and none for decreasing it. But, as we well know, the economy did not favor growth in the years that followed. Muzak filed for bankruptcy. During this same period, DMX similarly experienced a decline in its subscribers. It dropped from 84,000 customer locations in 2005 to just 70,000 locations in late 2009. As this brief discussion of the Muzak Agreement demonstrates, it is not a reliable benchmark for calculating an ASCAP/DMX license in 2010. DMX refused to enter a license premised on the formula created by the Muzak Agreement and, using hindsight, for good reason. Indeed, the competitive forces within this industry have continued unabated since 2005. Confronted with increased competition within the BG/FG music services industry, and with the economic pressures that cut across virtually every industry in the United States, DMX has had to substantially lower its prices to compete effectively. Even if ASCAP could show that a $41.21 per location rate was a reasonable rate for the year 2005, it has certainly failed to show that that would be a reasonable rate for any subsequent year. As a result, ASCAP has not shown that it can build a blanket rate for a multi-year license based on the $41.21 figure. Based on this analysis, it necessarily follows that the proposed per location rate of $49 for the year 2010, which ASCAP has indirectly derived from the $41.21 rate, is also unreasonable. ASCAP arrived at the $49 figure by dividing its proposed annual fee (originally calculated from the $41.21 figure) by the number of DMX locations in 2009. Since DMX had experienced a significant decline in its customer base, that simple division resulted in an increase in the per location rate between 2005 and 2009 from $41.21 to $49. ASCAP's proposed rate of $49 per location is over-reaching in the extreme. ASCAP and Muzak both expected that Muzak's per location rate in 2009 would be less than $41.21, not substantially higher. The only thing that produces a rate of $49 per location in 2009 is the unexpected decline in the number of locations for a BG/FG music service provider. A global economic decline of historic proportions is not a reliable basis from which to construct an increase in a licensing rate. In any event, Muzak entered the agreement with ASCAP in 2005 and accepted the risk of unforeseen market movements. DMX refused to execute a license with ASCAP then, and the decline in its number of *544 locations between 2005 and 2009 is now known. There is, therefore, no justification for imposing on DMX the consequences of a lost bet it did not make. ii. The Flat Fee for 2005 Through 2009 The second reason that ASCAP's alternative proposal is unreasonable is that it constructs a flat fee based on the $41.21 per location rate from 2005, and then applies that same flat fee for every succeeding year until 2009, despite significant fluctuations in the number of locations serviced by DMX. For example, DMX lost over 13,000 locations between 2005 and 2009. The cynicism behind this calculation is underscored by the switch that ASCAP proposes for 2010 and the years that follow. DMX recently entered an agreement with DIRECTV that increased the number of locations DMX serves by more than 20,000. Given this increase in locations, applying the flat fee to 2010 and the years that follow would reduce the per location rate below $41.21. As a consequence, and as just described, ASCAP simply abandons the flat fee when it no longer benefits from it and proposes a different method of calculating an annual fee for this latter period. iii. The Static Carve Out The third overarching reason that the ASCAP alternative is unreasonable is that it contemplates a static carve out. This proposal is flawed in several ways. For the years before 2010, the ASCAP credit formula only reduces DMX's license fee by the amount DMX has paid to its direct licensors; it does not allow DMX to reduce its payments to ASCAP based on the proportion of directly licensed music that it performs.[40] As for the three years between 2010 and 2012, a fixed annual credit is limited to the payments that DMX made in 2009 to its direct licensors. Thus, DMX can only reduce the cost of its public performance licensing fees by reducing the number of its direct licenses or cancelling its ASCAP blanket license altogether in the event it can rely solely on directly licensed music. Finally, the static carve out creates a windfall for ASCAP. These final two points warrant emphasis. The static carve out seeks to cabin the DMX direct licensing program by removing every incentive for DMX to continue to enter direct licenses. As noted, ASCAP proposes a fixed annual credit of $230,000 for each year between 2010 and 2012. This figure is derived from the amount DMX paid in 2009 to its direct licensors. The economic effect of this proposal is that DMX will be paying more in total licensing fees whenever it enters a new direct licensing agreement. DMX will have to continue to pay ASCAP its blanket licensing fee with no additional credit and will also have to pay the new licensor. This not only removes any incentive to continue with an expansion of the direct licensing program, it creates a strong disincentive to doing so. As Jaffe testified, this proposal is strongly anti-competitive. ASCAP economist Boyle similarly acknowledged that for DMX to benefit from ASCAP's alternative proposal, DMX would need to either reduce its reliance on directly licensed music so that its expenses dropped below the amount of the credit (and the administrative fee) or it would need to forgo an ASCAP blanket license altogether. ASCAP's proposal also creates a windfall for those ASCAP members who do not enter direct license agreements with DMX. Because (1) ASCAP is entitled to and does withhold from those members who enter *545 direct licensing agreements with DMX their share of ASCAP's revenue from DMX (no member is entitled to be paid by both DMX and ASCAP for the same licensing rights), and (2) DMX will only be credited for the direct licensing program as it existed as of 2009, with every additional direct license that DMX executes in 2010 through 2012, and with every expansion of DMX's reliance on directly licensed music, ASCAP will have extra income that it is not obligated to pass onto its members. If ASCAP uses that surplus income to increase its payments to those ASCAP publishers whose work is performed by DMX, but who have not entered into direct licenses with DMX—as ASCAP has indicated that it will do—those publishers will have even less of an incentive to join the DMX direct licensing program.[41] iv. The Administrative Charge of $25,000 Finally, ASCAP has not shown that its proposed charge of $25,000 per year for administrative costs is reasonable. ASCAP admits that it did no formal analysis or study to arrive at this figure. The $25,000 figure is based entirely on brief testimony from ASCAP's Chief Economist in which he describes the kind of work that would have to be done by ASCAP's repertory department and its economist's office to suspend payments to members who have joined DMX's direct licensing program. ASCAP has not supported this figure with any evidence of its actual costs. b. ASCAP's Defense of Its Alternative Proposal By the conclusion of the trial, ASCAP had largely abandoned any defense of its alternative fee proposal, its Chief Economist having admitted that it was inequitable to DMX.[42] Before abandoning the proposal, however, it principally made the following three arguments. It defended the $41.21 rate; it contended that the rate court has already ruled that a static carve-out license is entirely appropriate; and it urged DMX to apply for a per-segment license to avoid giving ASCAP a windfall in revenues. Not surprisingly, none of these arguments succeeds in salvaging ASCAP's alternative fee proposal. Nonetheless, these three arguments will be briefly addressed. i. The $41.21 Per Location Rate First, ASCAP defends the $41.21 per location rate as a reasonable base rate from which to calculate the fixed annual fee for the years 2005 through 2009, and from which to set, albeit indirectly, the annual fee to be paid for the period 2010 through 2012. Among other things, it contends that Music Choice and PlayNetwork agreed to pay the same rate; that the prevailing rates during the period between 1987 and 2004 were higher; and that ASCAP only agreed to accept a lower rate than had previously been paid because of the minimum guarantee contained in the Muzak Agreement. These arguments are unavailing. As has already been discussed, it is difficult to derive an effective per location rate from the Muzak Agreement. Indeed, the ASCAP and Music Choice negotiations for a final license fee illustrate the difficulties of *546 retrospectively applying the Muzak Agreement structure to determine final fees for a different BG/FG music service provider. In the Music Choice negotiations, the parties could not agree on how to apply the Muzak Agreement's 8% organic growth provision to a business that had almost doubled in size over the course of the license term. While ASCAP and DMX contest the effective per location rate that should be inferred from the Music Choice final license agreement, not even ASCAP contends that it is as high as $41.21. Moreover, it is undisputed that during the negotiations Music Choice rejected an ASCAP proposal that Music Choice pay a $41.21 per location rate throughout the license term. The 2009 PlayNetwork Agreement did adopt a $41.21 per location rate for the period between 2005 and 2009. Thus, ASCAP is correct that its recent agreement with PlayNetwork provides support for adoption of the $41.21 rate. That single agreement, however, is insufficient to support adoption of ASCAP's proposals, given the flaws in those proposals outlined above. Moreover, the adoption of the $41.21 rate in the 2009 PlayNetwork Agreement undermines ASCAP's position that it is entitled to receive a fee from DMX premised on a $49 per-location rate beginning in 2009. The second factor to which ASCAP points to support a $41.21 rate is the older industry agreements. While it is true that the fee based on the 5A locations under the 1987 and 1999 Form Agreements came from a higher per location rate than $41.21, by the time ASCAP and Muzak entered the Muzak Agreement in 2005, those earlier agreements no longer reflected an appropriate market rate. The record shows that the rates charged to the BG/FG music services industry have been declining for years, and the Muzak Agreement is consistent with this trend. Finally, the existence of a minimum guaranteed annual payment in the Muzak Agreement does not suggest that the $41.21 rate adopted in that agreement was artificially low.[43] It was Muzak and not ASCAP that requested the flat fee. Both parties expected that Muzak would increase in size and that the effective per location rate would continue to decline over the course of the Muzak license term. In short, Muzak would not have needed to agree to a guaranteed minimum payment in order to extract a lower price from ASCAP than that provided by the 1994 Form Agreement. ii. Muzak II Second, ASCAP contends that Judge Conner blessed the static carve out in his 2004 decision known as Muzak II. Responding to a request for clarification of Muzak I from Muzak and DMX Music, Inc., Judge Conner stated that: The Court need consider only those direct licensing relationships already in existence at the time of trial in determining a reasonable blanket licensing fee. Accordingly, the Order does not contemplate a blanket license fee mechanism that provides credits or discounts for direct licensing arrangement that applicants may enter into during the term of the license. *547 Muzak II, 323 F.Supp.2d at 590 (emphasis supplied). In so ruling, Judge Conner rejected the position taken by DOJ that the rate court could set a blanket license fee that was "open-ended" to accommodate growth in a direct licensing program. Id. at 591. The court recognized, however, that its view might nullify the benefits of direct licensing by requiring the music user to make double payments to ASCAP and new licensors. It endorsed ASCAP's solution to this problem, noting that ASCAP and the music users could simply enter into short-term blanket licenses. Id. at 592. Judge Conner did not have the benefit of the record created by this trial. When Muzak II was decided in 2004, DMX's predecessor had not begun a direct licensing program and Muzak had negotiated only two agreements, neither of which had yet taken effect.[44] Thus, the court was not presented with evidence of a robust direct license program like the one that DMX has since created. Moreover, no one in this current litigation has argued that it is wise or appropriate to order them to execute a short-term blanket license. Rate court litigation is expensive and burdensome. The parties' having made a significant investment in the litigation, they are entitled to finality for a reasonable period of time. Consequently, to the extent that the record permits a reasonable fee to be set for an adjustable fee license reflecting a dynamic direct licensing program, Muzak II will not be followed. See BMI/DMX, 726 F.Supp.2d at 355-56. iii. Per-Segment License Finally, ASCAP argues that many of the alleged infirmities in its alternative proposal, including the windfall it may reap, could be avoided if DMX merely applied for a per segment license. As already explained, DMX having requested a blanket license with a carve-out, it is entitled to receive a reasonable fee for such a license. ASCAP may not evade that obligation by assuring DMX that it could get a reasonable fee for a per-segment license. C. DMX's Proposal DMX proposes a mechanism for adjusting its blanket license fee in the years following 2009 to reflect the extent to which it relies on directly licensed music. Its proposal for the blanket license rate is the sum of two parts. The DMX proposal begins with a "floor fee" ("FF"). The floor fee is the minimum fee that DMX would pay for the benefits of an ASCAP license, even if all of the ASCAP music that DMX performs is covered by direct license agreements. It is intended to compensate ASCAP for the value it creates by constructing and administering the blanket license, i.e. aggregating the rights to publicly perform millions of musical compositions. The second component is the "unbundled music fee" ("UMF"), which it describes as the "pure" value of the performance rights for ASCAP music performed by DMX or its licensees. When combined, the unbundled music fee and the floor fee constitute the total "blanket fee" ("BF"), or the outermost limit of DMX's obligation to ASCAP. To give DMX a discount for its direct licensing program, the unbundled music fee is multiplied by the share of all DMX performances of ASCAP-affiliated music that is licensed to DMX solely by ASCAP ("Share Licensed via ASCAP"). This structure can be depicted algebraically as: Per-Location Fee=FF+[UMF × Share Licensed via ASCAP] *548 Thus, while DMX will always pay ASCAP the full amount of the floor fee, the amount of the unbundled music fee that ASCAP receives varies depending on the extent to which DMX subscribers play ASCAP works that are directly licensed.[45] If all of the ASCAP-affiliated music that DMX plays is directly licensed, DMX will pay ASCAP only the floor fee. If no music is directly licensed, then DMX will pay ASCAP the floor fee and the entirety of the unbundled music fee. Before describing this proposal in more detail, it is noteworthy that BMI did not contest that its licensing fee arrangement with DMX should be structured in a manner very similar to this. And, Judge Stanton recently approved this structure in setting the blanket license fee that DMX owes to BMI. BMI/DMX, 726 F.Supp.2d at 367. 1. The Unbundled Music Fee DMX identifies the appropriate benchmark for the unbundled music fee as the rate paid to those music publishers who have joined DMX's direct license program. Under the MCCL agreements, music publishers have granted DMX several rights in their musical compositions in exchange for a pro rata share of a $25 per location royalty pool. Like ASCAP, DMX estimates that approximately 10% of this royalty pool is attributable to the grant of mechanical rights as opposed to the public performance right. Thus, the royalty pool for public performance rights is limited to $22.50 per location. The $22.50 figure is further adjusted to reflect ASCAP's share of total performances on the DMX network. DMX concludes that 48% of its customer's performances were of works owned or controlled by ASCAP members.[46] Consequently, DMX proposes that the unbundled music fee should be $10.74 per location ($22.50 ×.48). This measurement includes all of the ASCAP-member music, whether or not DMX has obtained a direct license for the music. 2. The Floor Fee DMX calculates a floor fee of $3 per location. This represents the combination of $2.14 for BG/FG music service specific expenses and $0.86 in allocated general overhead expenses. Both of these numbers were derived from ASCAP's own records and the allocations that ASCAP makes of expenses associated with the BG/FG music services industry.[47] The first component—the industry-specific expenses—was calculated by dividing ASCAP's figure for its total costs tied to the licensing of the BG/FG music industry by the total number of locations served by that industry. The second component—a percentage of overhead to be applied against the industry revenues—was calculated by dividing (1) the costs ASCAP has allocated to the industry but that it has not *549 identified as directly tied to the licensing of the BG/FG music industry, by (2) the total revenue from the industry. This division results in a figure of 6.3% as representing the general overhead costs that should be allocated to the industry revenues. Because the floor fee component that is derived from overhead is a percentage of revenue, it will rise if the unbundled music fee increases. 3. The Blanket Fee: Sum of the Floor Fee and the Unbundled Music Fee Under the DMX proposal, the total blanket license fee is the sum of the floor fee and the unbundled music fee. If the Sony advance is not considered a royalty payment, the per location blanket fee is $13.74. 4. The Structure of the DMX Proposal is Reasonable and the Benchmarks that DMX Proposes are Appropriate DMX has shown that its proposal for an adjustable blanket license fee is reasonable. As an initial matter, isolating two components for the licensing fee—a floor fee and an unbundled music fee—is reasonable. It adequately compensates ASCAP for its service in providing a blanket license fee and provides ASCAP members with licensing income in the event they do not grant DMX a direct license. DMX has shown that a vibrant direct licensing program increases competition within this industry and should be permitted. The proposal acknowledges that ASCAP provides important services to both its members and to music users. Blanket licenses permit ASCAP members to receive a stream of income from their creations when they are performed in a broad array of outlets, including fora that individual members could not afford to police or license. And, blanket licenses protect music users against claims of copyright infringement. At the same time, the DMX proposal allows the appropriate incentives for DMX to continue and to expand its direct licensing program. DMX's methodology for calculating a floor fee and the calculations themselves appear to be reasonable and should be adopted. The fee is based on numbers taken from ASCAP's own books. It compensates ASCAP for both its direct costs from licensing the BG/FG music service industry and for a percentage of its general overhead costs. It is noteworthy that ASCAP has challenged neither the methodology nor the calculations associated with the proposed floor fee.[48] While ASCAP argued in summation that DMX's proposal does not adequately compensate it for the value it provides in aggregating individual compositions, it neither provided an alternative mechanism for such a valuation nor explained why the floor fee is insufficient to capture that value. There are two significant issues to address in order to determine whether the proposed benchmark for the second component of the DMX proposal—the unbundled music fee—is reasonable. The first is whether the $25 MCCL license fee is the appropriate benchmark for a per location licensing fee. The second is whether the unrecouped portion of the Sony advance should be included in the royalty income paid for licensing rights. a. The $25 Per Location Royalty Pool Over 850 music publishers and administrators have issued licenses to DMX since it began its direct licensing program in 2006. They have executed license agreements *550 that are relatively simple and unambiguous. Through these agreements they have granted mechanical rights and public performance rights to DMX in exchange for the commitment that they can share in a royalty pool premised on a rate of an annual fee of $25 per location. Among these licensors are prominent and sophisticated music publishers whose entire business is devoted to the licensing of musical compositions. These hundreds of agreements provide compelling evidence of the valuation of the right to publicly perform musical compositions within the BG/FG music service industry. As of today, 30% of DMX's musical performances are covered by direct licenses. DMX has shown that music publishers had several incentives to sign the direct licenses even though ASCAP had historically compensated them at a higher rate. First, DMX has been operating on an interim license since 2005. In the absence of a final license or a rate court determination, it was uncertain precisely what license fee ASCAP would be entitled to receive from DMX and to distribute to its members.[49] Moreover, DMX provided an incentive to licensors to be early adopters and to join its direct licensing program. It was in DMX's economic interest to maximize its use of music from direct licensees, and where it could, to decrease its reliance on the music from those who did not sign direct licenses. As a result, whatever fee a publisher might theoretically receive from ASCAP, if DMX substantially limited its use of that music, the publisher's income stream from ASCAP would necessarily decline. Conversely, by signing with DMX, a publisher could reasonably expect that its proportion of music performed by DMX and its licensees would increase. Thus, whether the royalty pool rate was judged as $25 or, as ASCAP emphasizes, just $10.74 for the ASCAP portion of the $25 fee, many music publishers decided that it was in their economic interest to become a direct licensor. The reasonableness of that judgment is confirmed by the experience of the major publishers. The percentage of DMX music performances controlled by the three major music publishers who did not sign direct licenses with DMX declined by roughly 20% between mid-2008 and the end of 2009. In contrast, Sony's share of DMX performances nearly doubled during that same period. The adoption of the $25 royalty pool rate is also consistent with an examination of the four factors that guide the selection of a benchmark. Music Choice IV, 426 F.3d at 95. While individual publishers are not "comparable" to ASCAP, taken as a community of 850 publishers and administrators, their collective decisions to execute direct licenses are comparable to the decision ASCAP makes in entering a license. The rights at issue are also sufficiently comparable. The MCCL includes a mechanical license and a somewhat broader public performance right, but the core of the license is the public performance right conveyed by the blanket license DMX seeks from ASCAP. The economic circumstances of the direct licensors and ASCAP are considerably different, but that difference is more than balanced by the degree of competition that the direct licenses inject into this marketplace. ASCAP has tried to show that DMX's direct licensors were misled or that they misunderstood the terms of the direct license and that once they became better informed, they declined to renew their direct *551 licenses. There is no evidence of any pattern of misrepresentations or omissions by either DMX or its agent MRI, and certainly none designed to deceive music publishers or administrators into signing direct licenses. The business of both MRI and DMX depends on their relationships and reputations within the music publishing industry. They have every incentive to protect those relationships. While there were a few email communications from MRI that were less than clear and testimony from several publishers that showed some confusion about certain license terms, ASCAP failed to demonstrate that there was any material misunderstanding about the direct licensing program. MRI vetted the direct license by negotiating it in the first instance with the most careful and scrupulous publishers. Where terms or language needed to be adjusted, it was. The license that has emerged from this process is clear and unambiguous in every term that is critical to the decisions at issue in this rate court proceeding. As for the few publishers that did not renew their licenses, their decisions do not undercut the stronger evidence represented by the many other publishers and administrators who did renew their direct licenses. Taking a broader look at the license fee, recent history confirms that a $25 fee is within the range of reasonable rates. The overall trend in the rates for licensing within the BG/FG music services industry has been a declining one. In the last four years, the industry has experienced enormous competitive pressures from within and from without. DMX's per location revenues have consistently declined. A royalty pool rate of $25 represents 5.6% of DMX's music service related revenue. When the floor fee and ASCAP unbundled music fee are combined, they total $13.74 and constitute 3.1% of DMX's music service related revenue. ASCAP has not shown that a fee set at 3.1% of such revenue is unreasonable, even when applied to a service almost entirely dependent on the distribution and licensing of "pure" music. b. The Sony Signal The only remaining issue of serious concern is the effect of the Sony advance on this analysis. It is undisputed that the "Sony signal" had an enormous impact on DMX's direct licensing program. Signing a major music publisher was necessary for DMX to achieve the growth in its direct licensing program that it has in just a few years time. DMX also acknowledges that, to the extent that the Sony advance is determined to be a royalty payment, the effective royalty pool under the MCCL agreements should be increased by the percentage of the advance that will be unrecouped. Under this approach, the unbundled music fee could rise from $10.74 to as high as $17.92.[50] DMX has also shown, however, that a significant and in all likelihood partially unrecoupable advance was necessary to induce the first major music publisher to *552 enter into an MCCL. Indeed, BMI paid Universal a sizeable advance to stop Universal from signing a direct license. DMX was willing to pay this premium because it needed at least one major music publisher to participate in its program. Independent music publishers wanted to know if a major publisher had executed a direct licensing agreement; and the direct licensing program increased exponentially after DMX was able to tell publishers that Sony had indeed joined the program. In the recent rate court proceeding between DMX and BMI, Judge Stanton found that the unrecouped portion of the Sony advance was "reasonably viewed as a cost of entry into the market, rather than as allocable to royalties." BMI/DMX, 726 F.Supp.2d at 361. That same finding is appropriate here. This analysis is confirmed by the fact that the Sony agreement with DMX included a confidentiality provision that prevented the market from learning that Sony had been given an advance, much less an unrecoupable advance or the size of that advance. Thus, independent publishers executed direct licenses at the $25 royalty pool rate with no basis to believe and no expectation that the pool could be enlarged. Moreover, there is no evidence that since the existence of the Sony advance became public at the BMI/DMI trial in February 2010 and through the issuance of Judge Stanton's opinion in July 2010, any direct licensor has protested to DMX that the $25 pool rate should be increased to include that advance and no one has asked for a similar advance.[51] In their closing arguments, both parties requested that this Court determine what a reasonable fee is for a DMX blanket license based on the record created by this trial, even if it means adopting a fee or fee structure that varies from that adopted by Judge Stanton following the rate court proceeding earlier this year between DMX and BMI. DMX has shown at this trial that a royalty pool set at a rate of $25, and an adjustable blanket license fee for ASCAP premised on a combination of a floor fee and an unbundled music fee, as computed through DMX's proposal, is reasonable. As it turns out, Judge Stanton also approved a similar fee structure and the $25 royalty pool rate. His independent judgment on nearly identical issues presented in the context of the BMI and DMX litigation serves only to confirm the reasonableness of DMX's proposal here. CONCLUSION DMX's proposal for a blanket license fee with adjustments for its ongoing direct licensing program is adopted. SO ORDERED. NOTES [1] By the terms of the consent decree, the matter was to be ready for trial within one year of that date, subject to an extension of no more than one year, absent good cause for a longer extension. AFJ2 § IX(E). ASCAP's application was reassigned to this Court on July 22, 2009. DMX and ASCAP reached agreement in September 2009 on interim fees to be paid pending the outcome of the rate court litigation. Pursuant to scheduling orders of August 21, 2009, March 1, 2010, and May 24, 2010, the parties completed fact and then expert discovery, and a trial was scheduled for the fall of 2010. [2] Greene is a Professor at the Stern School of Business, New York University, where he holds the Toyota Motor Corporation Chair in Economics. He researches and teaches in the area of econometric theory and methods, including the application of these methods to several industries. He also has written and taught on the microeconomics of the entertainment and media industries. Greene testified on behalf of ASCAP in its 2007 rate court proceedings against AOL, RealNetworks, and Yahoo!. [3] Jaffe is the Fred C. Hecht Professor in Economics and Dean of Arts and Sciences at Brandeis University. Jaffe's course offerings include classes in microeconomics, industrial organization, and antitrust and regulatory economics. He has served as an expert on the valuation of intellectual property in numerous proceedings and was an expert on behalf of DMX in its rate court proceeding against Broadcast Music, Inc. ("BMI") earlier this year. [4] Candell is a consulting economist and Senior Vice President in the Boston office of Compass Lexecon, an economic consulting firm, where she focuses on antitrust, regulation, and other applications of microeconomics. She has worked on numerous matters involving the valuation of intellectual property, including the public performance of music by major media users. Candell served as an expert on behalf of DMX in its rate court proceeding against BMI earlier this year. [5] Background music, as the term suggests, plays quietly in the background of a space, be it an elevator or a large commercial space. Foreground music plays loudly enough to be a significant part of the experience of the space. Clothing stores for young consumers are typical users of foreground music. [6] The five industry leaders are DMX, Muzak LLC ("Muzak"), Music Choice, Inc. ("Music Choice"), PlayNetwork, Inc. ("PlayNetwork"), and Trusonic, Inc. ("Trusonic"). [7] The DMX programming department suggests to customers the channels it believes are best suited to a given business's purpose, and often customers follow these suggestions. DMX does not, however, systematically track which channels are actually played. [8] Disparities between the off-premise channels and the on-premise programming arise when DMX has acquired the public performance right to a particular work, but not the accompanying reproduction right. Since providing a composition to on-premise customers involves reproducing the work, only compositions for which DMX holds both the performance and reproduction rights can be used in on-premise programming. ASCAP works constitute a larger percentage of on-site programming than off-site programming; ASCAP estimates that its compositions constitute 41% of DMX's off-premise performances and about half of DMX's on-premise performances. [9] Pursuant to ¶ 5(d) of the 1987 Form License, "gross billings" included all money the BG/FG service collected from subscribers, including revenue derived from the lease of equipment or installations relating to the service. [10] Per-segment licenses are described infra. [11] Muzak offers scores of music channels. An ASCAP survey of two of them showed that 58% of their music in the five year period ending in 2009 came from the ASCAP repertory. [12] While one ASCAP witness testified at trial that, during the negotiations, he personally believed that Muzak's business would contract due to the stiff competition it faced and its out-moded business model, that testimony is less than credible. Although a highly placed ASCAP official, he shared that view with none of his colleagues, and internal ASCAP documents and ASCAP's negotiation strategy tell a different story. [13] There are roughly [REDACTED] Muzak affiliated companies that ASCAP separately licenses. These franchises are currently paying ASCAP interim fees of $41.14 per location per year. [14] ASCAP took the position that if a music user's growth exceeded the 8% growth limit, the base license fee for the succeeding year should be increased. Music Choice insisted that the base fee remain static at the rate set for the first year of the license term, regardless of whether it exceeded the 8% organic growth limit. [15] Prospectively, the 2010 Music Choice Agreement set an interim rate of $41.14 per location per year. [16] For the period beginning January 1, 2010, the parties agreed to an annual interim rate of $43.00 per location. [17] This figure is the number of locations that Trusonic reports to ASCAP and has not been independently verified by ASCAP. [18] DMX, like all businesses, has also been affected by the general economic downturn. The decline in the economy has forced DMX customers to close store locations. DMX's total customer store location count has recently increased, however, due to DMX's 2009 agreement DIRECTV. [19] In 2001, it was common for major clients to pay DMX between $50 and $70 per location, each month. Currently, the highest monthly fee on the DMX "rate card" for music services is $42. The rates DMX has proposed in recent competitive bids indicate that the rates for relatively large customers may be significantly lower than the current rate card suggests. DMX recently proposed a rate of [REDACTED] per month for a 400-location retailer, [REDACTED] per month for a 900-location restaurant chain, and [REDACTED] per month for a 2,700-location retailer. [20] In June 2005, DMX was paying ASCAP annual interim fees of $49.50 per customer location and it was accruing royalties to be paid to BMI at the rate of $36.36 per location per year, the rate BMI was seeking in its license negotiations with DMX. [21] DMX also promised licensors faster payments and more detailed reporting of the music performed than they received from PROs. [22] DMX occasionally removes lyrics that its customers deem offensive. [23] In early 2010, DMX amended the MCCL scope of rights provision to include music video and in-flight services. In addition to those publishers who have signed MCCL agreements since the inclusion of this new provision, eleven publishers that had signed the original MCCL have now signed renewal agreements that include this term. [24] Publishers regularly have the obligation under "mechanical" licenses to distribute half of the money they receive from licensees to songwriters. [25] In 2006, DMX signed 65 MCCLs. In 2007, it signed 50 more. In 2008, it concluded 150 agreements. In 2009, it added 300 licenses. As of September 2010, it had signed 250 new agreements in 2010. [26] For example, DMX has directly licensed songs by Frank Sinatra (Barton Music), Rascal Flatts (WAMA), TLC (The Royalty Network), Elvis Presley (HoriPro Entertainment), David Bowie, Elton John (Realsongs), Britney Spears (Realsongs), Barbara Streisand (Realsongs), Julio Iglesias (Realsongs), Kenny G. (Realsongs), Aerosmith (Stage Three Music), Roy Orbison (Integrated Copyright Group), Etta James (Arc Music), B.B. King (Arc Music), Yerba Buena (Shelly Bay Music), Whitney Houston (Lastrada Entertainment), Mötley Crüe (Downtown Music Publishing), Sheryl Crow (S1 Songs America), Richard Rodgers and Oscar Hammerstein (Williamson Music), Irving Berlin (Williamson Music), Ray Charles (Sure Fire Music), Dolly Parton (Sure Fire Music), Buddy Holly (MPL Publishing), Paul McCartney (MPL Publishing), Jelly Roll Morton (MPL Publishing), Justin Timberlake (Reservoir Media Management), and Enrique Iglesias (Reservoir Media Management). [27] Fifty-four publishers have signed amendments that renew the license for an additional number of years. Twenty-four have renewed subject to automatic renewal provisions. [28] Levin did not renew her direct license with DMX after BMI contacted her. She testified for BMI at its 2010 rate court proceeding against DMX. Levin testified that she understood the MCCL to convey only the "mechanical" right and not the public performance right. That testimony is simply not credible, given the unambiguous terms of the MCCL. Levin has a law degree, has practiced entertainment law and would be expected to understand unambiguous contract terms. She also admitted that she had received an email from DMX that stated "DMX ... is seeking a catalog direct license with publishers including all necessary performance and pre-production rights in order to distribute the DMX service to business establishments in the U.S., regardless of the delivery methodology." (Emphasis supplied). When questioned about this document, Levin asserted that DMX was "hiding" or "obscur[ing]" the fact that it was seeking to engage in direct licensing to acquire performance rights. [29] DMX claims that it does not intend to enter into direct license agreements with all four major publishers. DMX explains that its plan is, having secured licenses from one or two majors, to "round out" its programming with the direct licensing of independent publishers. [30] The chairman and CEO of each of the four major music publishers is a member of ASCAP's board of directors. Of the twenty-four seats on ASCAP's board, twelve are held by music publishers. [31] Negotiations with Sony began at some point in 2006. The Sony catalog includes works performed by many significant artists, including the Beatles, Elvis Presley, Jimi Hendrix, Joni Mitchell, Taylor Swift, and Lady Gaga. [32] ASCAP suspended its distributions to Sony in March 2008 based on the Sony/DMX direct license. [33] Sony was apparently unaware of DMX's "misreading" of the BMI royalty report during the negotiations leading up to the 2007 MCCL. [34] Sony believed an extension was appropriate as a means of compensating DMX for Sony's delay in delivering to DMX the song information necessary for DMX to start increasing its use of the Sony repertory and to maintain an amicable relationship with DMX. [35] Judge Stanton rejected BMI's license with Muzak and its other recent licenses within the BG/FG music services industry as benchmarks. Id. at 357-59. [36] Shenandoah Valley addressed the AFJ, the relevant portion of which (§ IX) is identical to the same-numbered provision of AFJ2. [37] The AFJ2, its predecessor—the AFJ—and the BMI Consent Decree all provide that when a PRO receives a request for a "license for the right of public performance of any, some or all of the" works in its repertory, it shall "advise" the applicant "in writing of the fee" it "deems reasonable for the license requested." AFJ § IX(A); AFJ2 § IX(A); BMI Consent Decree § XIV(A). [38] At the August 20 conference, ASCAP's principal trial counsel argued that "what Judge Conner has ruled is that ASCAP must take account of these direct licenses that exist at the time of trial. ASCAP is not required to offer a license that flexibly adjusts as time goes on and as they enter into more direct licenses." Later, at the same conference, he acknowledged that "[w]e will have to take account of their direct licensing in formulating our proposal." [39] DMX does not dispute that 10% of the $25-per-location pool is properly attributable to the grant of mechanical rights. [40] Of course, DMX must always pay an administrative fee to ASCAP as well, and therefore is necessarily out-of-pocket both the administrative fee and all of the expenses DMX incurs in running the direct licensing program. [41] The premium that ASCAP would be able to give such members currently is roughly estimated to run as high as 40% above the distributions that they would otherwise receive from ASCAP. If DMX increases the percentage of its performances that are directly licensed or if ASCAP reduces its administrative expenses, the windfall to ASCAP members who do not enter MCCLs could be even higher. [42] ASCAP's summation did not even address its alternative fee proposal or argue that it should be adopted as reasonable. [43] While ASCAP has argued that it only consented to a lower rate than that contained in the 1994 Form Agreement because Muzak agreed to give ASCAP a guaranteed minimum, there is no credible basis to find that that is so. Among the advantages to ASCAP of a minimum annual payment were the reduced need to conduct audits of Muzak's business, the existence of a steady income stream, and the opportunity for increased efficiency in making distributions to its members. [44] After Muzak II was issued, Muzak apparently abandoned its direct licensing program and settled with ASCAP. [45] An alternative way to reach the same result is to subtract from the blanket fee the amount of the unbundled music fee that should be attributed to the direct licensors. This formulation is useful when comparing the DMX proposal to the ASCAP proposals. Algebraically, this can be depicted as: Per-Location Fee=BF-[(BF-FF) × (1-Share Licensed via ASCAP)] [46] DMX calculates the percentage of ASCAP music from ASCAP's royalty distribution data from the period beginning in the third quarter of 2005 and ending in the fourth quarter of 2008. This figure is "share-weighted," meaning that it accounts for the fact that certain works are only partially controlled by ASCAP. [47] In 2009, ASCAP allocated 10.82% of its revenues from the BG/FG music services industry to its total costs for administering the licenses in that industry and to an allocated share of general overhead expenses. [48] In contrast to the recent BMI litigation with DMX, in which BMI argued for alternative measurements of a floor fee, BMI/DMX, 726 F.Supp.2d at 361-64, ASCAP did not cross-examine DMX's experts regarding the floor fee. [49] For this among other reasons, ASCAP's argument that it has been distributing more money to its members for the public performance of their music by DMX and its licensees than those members have received from DMX under their direct licenses for the very same performances does not persuade this Court that the $25 royalty pool rate is too low. [50] Assuming that any unrecouped portion of the Sony advance constitutes a royalty payment, determining an effective rate for Sony is difficult because it requires making assumptions about the extent to which DMX will recoup the outstanding balance of the advance before the license term expires. If DMX continues to use Sony music at the level it did in the first quarter of 2010(12%) and the number of DMX locations remains stable (94,896), the effective per-location rate for the Sony license is $25.53. This figure also assumes that 10% of the Sony advance is attributable to the reproduction right as opposed to the public performance right. Next, to derive an unbundled music fee based on DMX's entire direct licensing program, the rate paid to Sony must be balanced with the effective rate paid to all other direct licensors. This is achieved by taking a weighted average of the effective rate paid to Sony ($25.53) as opposed to the rate paid to all other publishers ($10.74). The average is weighted based on the respective share of Sony performances as opposed to non-Sony directly licensed performances over the life of the Sony contract. Using figures from the first quarter of 2010, when Sony compositions constituted 12% of DMX's performances and non-Sony directly licensed works were 13% of DMX's performances, the result of this calculation is $17.92. Finally, it is worth noting that any increase in the unbundled music fee also causes an increase in the component of the floor fee that compensates ASCAP for its overhead expenses. Thus, with an unbundled music fee of $17.92, the total per location blanket license fee rises from $13.74 to $21.39. [51] The lack of a reaction by independent music publishers to the news of the Sony advance is not altogether surprising. The four majors are called majors for a reason. They hold a disproportionate amount of power in the industry due to the size and quality of their catalogs. The fact that they would be able to command a premium to join a novel program would not be startling.