right. First, if the subsequent party is a bona fide purchaser or licensee, it would need only look for a filing with the Copyright Office. If a security interest in the copyright is not registered at the Copyright Office, the purchaser or licensee would take free and clear of the security interest. Second, if an outright transferee elected to conduct a search for UCC recorded security interests, the search would only need to be in one state, namely the state where the debtor is located.

REAL WORLD EXAMPLE OF THE PROBLEM OF PEREGRINE

I would now like to share with the Subcommittee a recent true-life example of the problems that lenders and borrowers are facing as a result of Peregrine and Avalon. A CFA member financial institution in Chicago recently desired to make a bridge loan to a software developer who had begun developing a software program. This software program was the developer's primary asset and constituted most of the collateral for the proposed loan. The bridge loan was to be the beginning of what both the lender and the software developer expected to be a significant business relationship.

The lender was advised that under Avalon, it could not perfect its security interest in the software program of the developer until the program was registered with the Copyright Office. It required several weeks for the lender to convince the developer to register the software and prepare the necessary papers. Both the lender and the developer were quite concerned that the confidential nature of the software's source code could be compromised by registering it in the Copyright Office. In addition, the lender was concerned that there might be insufficient code developed to justify filing in the Copyright Office. The lender was also troubled that it might need to update the registration at frequent intervals, and also update its security interest, to reflect changes in the program or to reflect new programs developed by the borrower.

In the end, because of Peregrine and Avalon the lender was forced to slow down the transaction and require the software developer to register its software program, something the developer did not want to do. The lender now has a significant obligation to monitor the developer's ongoing software development work to determine whether enhancements or modifications to the registered software require additional registrations and security interest filings with the Copyright Office. In the secured financing arena, such situations do not enhance relationships between borrowers and lenders or promote efficient financing transactions.

CONCLUDING REMARKS

On behalf of over 300 members institutions of the CFA, I want to thank the Chairman for holding this hearing on this important issue that greatly affects the asset-based lending industry. As a result of the fundamental differences and difficulties in perfecting a security interest under the Copyright Act as opposed to the UCC, borrowers and their lenders have had to incur significant cost, delay, and administrative burden, or in some situations, they have had to forgo otherwise attractive financing opportunities altogether. As the software and related information technology industries continue to grow at a dazzling pace, the problems associated with financing copyrighted and copyrightable material under current federal law will only increasingly complicate or preclude credit extensions to such businesses. CFA believes that its limited amendment to the Copyright Act offers a focused solution to a significant problem. It will allow a secured lender with a UCC perfected security interest to prevail over bankruptcy trustees and other secured parties when copyrighted and copyrightable material are offered as collateral. Such a modest change in the law to address a currently untenable situation will enable asset-based lenders to more efficiently and effectively fulfill their role in providing working capital to the small and medium sized businesses of America. It will also conform the law covering the treatment of copyright security interests in bankruptcy to the law covering patents and trademarks in that arena.

The CFA bill is not in derogation of any attempt to establish a comprehensive federal system for the recordation of security interests in intellectual property generally (including copyrights). It is an interim measure to correct an immediate problem that has adversely affected the secured financing of businesses that need to pledge copyrighted or copyrightable material as collateral, and as a measure to level the playing field for the treatment of all security interests in intellectual property in bankruptcy.

For all practical purposes, the CFA bill will simply allow a secured lender, who has loaned money to a business secured by copyrighted and copyrightable assets of the business, to establish priority over a bankruptcy trustee through a filing under the UCC-as can be done with patents and trademarks. This way, in the unfortu

nate event that the business borrower seeks bankruptcy protection and the copyright assets are sold, the lender, who supported the business by extending it credit, will be paid ahead of the bankruptcy trustee, as intended by commercial and bankruptcy law, and will not lose out because of the problems in perfecting a security interest under copyright law.

Mr. COBLE. Mr. Brennan.

STATEMENT OF LORIN BRENNAN, ESQ., EXECUTIVE DIRECTOR, GRAY MATTER, LLC, ON BEHALF OF THE AMERICAN FILM MARKETING ASSOCIATION

Mr. BRENNAN. Thank you, Mr. Chairman, I am here on behalf of the independent sector of the motion picture business. We have 140 companies engaged in producing independent pictures as well as 22 major financial institutions engaged in the business of copyright lending. I hear that some of my member institutions are the entertainment divisions of people who are also members of my colleagues' associations, so this will be interesting.

I am here to give our strongest opposition to the mixed filing system proposed in the ABA proposal and the CFA proposal. We opposed this in 1993 and we oppose it now because it would devastate lending in our business, raise costs to simply unacceptable levels. We represent the lenders who make loans exactly like were made in Peregrine. Our lenders find that the Peregrine decision actually affirmed their longstanding practice of making loans-and speaking as a lawyer who has actually made loans and documented and administered loans exactly like that in Peregrine, I would suggest that the lender in Peregrine got into trouble because it simply did not follow the ordinary standards of care by lenders who actually make these types of loans. That doesn't mean the system is perfect and we will propose some changes, but I would like to give you three basic ideas.

First, I wanted to explain why the mixed filing system will not accomplish its goals; second, why it will drive our cost to excessive levels; and third, a proposed solution.

First, why won't the mixed filing system accomplish its goals? Priority doesn't matter. The real issue is what happens when the secured creditor forecloses? Imagine this situation. Day one, lender files a mortgage of the copyright and files it only at the State level. One month later we have an exclusive licensee who records at the Federal level. One month later the secured creditor forecloses. Who owns the copyright?

Under State UCC, the lender becomes the transferee and owns the copyright free of the exclusive license; but under Federal law, the Federal transferee, since he recorded, federally owns the exclusive license. This is a direct conflict between State and Federal law. It cannot be resolved by any finessing of drafting. Only one can prevail.

Now, under both of the proposals we hear today, it looks like the idea is that the secured creditor would lose. That means a subsequent transferee would still prevail. But what does that do to the lender's collateral value? That means if the lender does not also record federally, it has no value. Mr. Chairman, it does no good for a bank to foreclose a loan on my car if my brother-in-law can still drive it. This statute will not eliminate at all the need to file and search federally. If Federal laws prevail, we must do that.

Number two, let's look at the cost. The problem with copyrighted works like motion pictures is we have very complicated chains of titles with many transfers. We have to look up a lengthy chain of title. I have included in my testimony an example of a motion picture and what it would take to finance.

Let's assume we were going to make a new movie based on the Terminator movies, Terminator 3. Now, there are prior works here. There are prior screenplays and two prior pictures. I have included a copy of a copyright report that we would see but I have excluded all the secured creditors. There are 94 transferees in the prior chain of title. We would have to locate them to see whether or not they have filed a security interest. Yes, we only have to search in one State but we don't know where they are. To find all 94 transferees in all 50 States would cost us more than $30,000. Once we then find the transferees, we have to order a search. That is 94 companies at $34 a search. Another $8,000.

Then we get the UCC-1s but the UCC-1s are not related to Terminator; they are related to a collage of people. We will have thousands of UCC-1s that we have to order copies from and to read for more money. That doesn't include the individuals. You can't search by individuals by country. So the cost here in the chart I have given you, it will cost us almost $60,000 to do what now costs us $250 under the current system and we can't even be sure we found everything.

That doesn't mean we think the current system is perfect. We have suggested we need to deal with after-acquired property and floating liens but they should be done in the Copyright Office. Copyrights are Federal works. They are supported federally. We need a Federal system. We have given a proposal in here of one way to do it in which we give constructive notice to documents filed in the Copyright Office against the transferee but we support the goal of improving the law in order to allow after-acquired properties and floating liens. Copyrights are national works. They can only be supported by a national register and therefore this needs to be done in the Copyright Office. We cannot live with the mixed systems proposed by either the ABA or CFA proposal. Thank you. [The prepared statement of Mr. Brennan follows:]

PREPARED STATEMENT OF LORIN BRENNAN, ESQ., EXECUTIVE DIRECTOR, GRAY MATTER, LLC, ON BEHALF OF THE AMERICAN FILM MARKETING ASSOCIATION

INTRODUCTION

My name is Lorin Brennan. I am appearing as Special Counsel for AFMA (the American Film Marketing Association) and their Affiliated Financial Institutions (AFIs). AFMA is a trade association for 142 independent (non-studio) motion picture and television production and distribution companies. The AFIs, a division of AFMA, consist of 22 major banks and financial institutions. Neither I nor AFMA have received any federal grant, contract or subcontract in the current or preceding two fiscal years.

All of our member companies are familiar with secured copyright lending and the Peregrine decision (In re Peregrine Entertainment, Ltd., 116 B.R. 194 (C.D. Cal. 1990)). On their behalf, I am here to register our strongest opposition to the proposed Federal Intellectual Property Security Act (FIPSA). We oppose FISPA be

cause:

• FIPSA will decimate the ability of motion picture producers to finance their productions, threatening tens of thousands of jobs.

• FIPSA will skyrocket the risks and legal uncertainties of secured copyright lending to ruinous levels.

• FIPSA will not and cannot accomplish its stated objectives, subjecting copyright borrowers and lenders to unacceptable risks for no benefits.

In my brief remarks I will identify the reasons why we have come to these conclusions. They are not new. We have opposed the FIPSA "mixed filing" scheme for years, and communicated our opposition to Congress, the Copyright Office and to the American Bar Association (ABA).

At the same time, we acknowledge that the current system of secured copyright financing needs improvement, especially for "floating liens" and "after-acquired property." We have advocated the need for constructive change to the Administration, the Copyright Office, and now Congress. We support the goal of FIPSA, but cannot in any way support its methods.

In have attached to this statement copies of some of our extensive work in this regard. They describe the many reasons why we cannot support the "mixed filing" system in FISPA. They also contain our suggestion for a workable solution.

WHY FIPSA WILL NOT WORK

Copyrights exist solely by reason of federal law. They are intangible assets, simultaneously everywhere, and highly divisible. They often have complex chains of title. As national assets, they are supported by a single, unified federal recording system in the Copyright Office. This system indexes transfers of copyright ownership against registered works.

Article 9 of the Uniform Commercial Code is state law. It deals with assets that are either located in any easily identifiable place, e.g. hard goods, or, in the case of intangibles, presumed to exist only at the location of the debtor. They rarely have complex chains of title. Article 9 indexes security interests against the debtor in numerous filing systems. As one commentator puts it: "Variations from state to state are legion; some are authorized by alternative versions of the [Uniform Commercial] Code itself; others are local frolics." (Barkley Clark, THE LAW OF SECURED TRANSACTIONS UNDER THE UCC, ¶2.12[1] (1994 Rev. Cum. Ed.).)

The collateral, focus and methodologies of the copyright system and the Article 9 system are marked opposites. Mixing them is a recipe for disaster.

1. FISPA Will Not Eliminate The Need To Make Dual Filings

The proponents of FIPSA complain that the decision in Peregrine requires a creditor loaning against a copyrighted work to search and file in the Copyright Office in addition to filing against the other assets under Article 9.

Whether or not this is a burden, FISPA will not eliminate it. To see why we need only ask: What happens when the secured lender forecloses? Consider the following (ignoring the 30-60 filing windows in the Copyright Act):

May 1: Copyright owner grants a security interest in a registered copyright to a Lender who records under UCC.

June 1: Copyright owner grants an exclusive license to a Licensee who records in Copyright Office.

July 1: Lender forecloses and becomes a transferee at foreclosure sale.

Who own the copyright, the Lender or the Licensee?

This example illustrates the conflict between the exclusive Licensee who records in the Copyright Office under federal law, and the Lender/Transferee at a foreclosure sale under state law. The conflict is not just in the priority schemes. It is in the basic system for transfer of ownership of the copyrighted work.

Under Article 9-105(d), a "debtor" includes a later owner of collateral, and Article 9-306(2) continues a security interest in collateral “notwithstanding sale, exchange or other disposition thereof" unless the lender releases. (Barkley Clark, THE LAW OF SECURED TRANSACTIONS UNDER THE UCC, ¶¶1.02[2], 2.11[1][a] (1994 Rev. Cum. Ed.).) Article 9-504(4) provides that "[w]hen collateral is disposed of by a secured party after default, the disposition transfers to the purchaser for value all of the debtor's rights therein, discharges the security interest under which it was made, and any security interest or lien subordinate thereto." In other words, per Article 9 the Lender takes the copyright free of the Licensee's later recorded interest.

Under Section 205(d) of the Copyright Act, however, "[a]s between two conflicting transfers,... the later transfer prevails if recorded first" in the Copyright Office in the manner necessary to give constructive notice. Under this section, since the Licensee recorded federally and the Lender did not, the Licensee prevails over the Lender.

This is an irreconcilable conflict between the ownership transfer provisions of state and federal law. No amount of drafting can finesse it. Only one system can prevail.

Although not entirely clear, FIPSA apparently does not intend to eliminate federal preemption of the ownership transfer rules for copyrights. Thus, under FIPSA the Licensee would still prevail.

But this means the Lender's security interest has no value against subsequent assignees or exclusive licensees unless the Lender also records in the Copyright Office. Even under FISPA, any prudent lender must still search and record in the Copyright Office to ensure the continuing value of its security.

Thus, FIPSA does not and can not eliminate the "dual filing burden."

2. FISPA Will Significantly Increase the Costs Of Copyright Lending

FISPA provides that the relative priority between secured lenders and lien creditors is determined solely by state law. Copyrighted works often have complex chains of title, with many tiers of derivatives works, licenses, sublicenses and sub-sublicenses. Under Article 9, foreclosure by a senior secured creditor anywhere higher up in the "chain" wipes out a junior interest. This makes its critical for lender to find any senior interests before it makes a loan.

Under current law, the lender can conduct a single search of the records in the Copyright Office to find all prior copyright liens. Under FISPA, a lender will now need to search the UCC filing systems maintained in the fifty states. This burden will be enormous.

Let me illustrate the magnitude of the problem with a practical example. Assume an independent producer wants to finance a new movie, Terminator 3, based on the popular Terminator and Terminator 2 pictures. Since Terminator 3 will be a derived from several prior works (screenplays and movies), the production lender certainly wants to know all prior security interests against any of these prior works since they would be superior to its loan. Attached is a Copyright Report for the Terminator pictures showing the recorded chain of title, but with the copyright mortgages omitted. If FISPA passes, this is what secured lenders will see in the future. Let us try to find all prior copyright mortgages using only this report.

The report shows numerous prior transfers. Any transferor or transferee could have granted a security interest in its rights so we must search each one. Ignoring duplicate entries and transfers that may not affect our new production, it looks like there are still 94 separate entities to search. The Copyright Report does not show where they are located, so we must search under the rules in all fifty states and the District of Columbia.

There are several state variants regarding the place to file a security interest. For a general intangible like a copyright, 26 states only require filing in the Secretary of State, while 14 others can also require an additional filing in the county where the debtor is located. (HAWKLAND, LORD & LEWIS UCC SERIES, 889-401 et. seq. (1997 ed.).) According to one major searching service, there are 6,400 potential filing jurisdictions in the United States. How will our production lender find all prior lien filings, and how much will it cost?

Here is an analysis based on the going charges of a well known search company. • Locating the Corporations: It looks like 90 of the entries in the Chain of a Title are corporations. We can conduct a 50 state search to locate their place of incorporation at $365 per debtor, i.e. $365 x 90 = $32,850.

• Searching the Corporations: Now that we know where they are located, we must search the local UCC filings there. Searching fees vary per state. In California, a search at the Secretary of State costs $34. No county search is required for intangibles. If all of these companies are headquartered in California, then we must conduct 90 UCC searches at a cost of $34 x 90 = $3,060. Copies of Filing: The UCC searches only disclose the UCC-1s filed against the debtor, not against the copyright in Terminator. One transferor alone, Carolco Pictures, had more than 1,000 UCC-1 filed against it. (Each time it produced a picture, the lenders filed a UCC-1 to cover the physical materials). To find out which UCC-1s apply to Terminator, we must order a copy of each one. The cost in California is $1.25 per page. If we make the conservative assumptions that each UCC-1 is only 1 page, and that all 90 debtors only have only 2,400 UCC-1s in total filed against them, this means an additional cost of $1.25 x 2,400 = $3,000.

• Reading The UCC-1s: Then somebody has to read all 2,400 UCC-1s to determine which ones apply to Terminator. If a legal professional could read one UCC-1 a minute, it would still take 40 hours to read them all. If were only charged $100 per hour, that is an additional $4,000.