To: Jeff Hunter, Director, Office of Debt Management From: Drew Forbes, Treasury Desk, Daiwa Securities America Eric Welles, Repo Desk, Daiwa Securities America Re: Securities Lending Facility Date: August 10, 2006 We strongly support the adoption of a securities lending facility at the Treasury Department. We agree with the view that such a facility would minimize or eliminate problems like those in 2003, thereby enhancing the efficiency and liquidity of the Treasury securities market. Your counterargument about promoting moral hazard and preempting private market solutions is well taken, but we suspect that usage of the facility would be so limited that these would not be problems in practice. The system described in your report is generally sound. We would like to highlight two aspects that we view as essential ingredients of the final program.

First, the system should be of the fixed-rate, variable-quantity type. This is the only type of system that will adequately address issue shortages that might arise. Second, the implied repo rate should be zero percent.

This will discourage unnecessary usage and make the system a backstop rather than frequently used borrowing facility. We recommend a change to another aspect of the system described in your memo. Specifically, we feel that forward settlement and term transactions would discourage usage and leave the system ineffective in countering many market disruptions. We feel the system should offer same-day settlement and overnight lending as the core transaction. Such a system would offer maximum flexibility and therefore be most effective in remedying market imbalances. Some might argue that same-day settlement and overnight borrowing might lead to abuse of the system or gaming by market participants. However, we do not feel this will occur, as the zero percent repo rate would discourage overuse and abuse.

Moreover, if the Treasury detected abuse, the problem could be corrected with a verbal reprimand to the violating dealers and the imposition of administered limits on their access to the facility. We hope you find our comments useful. Please call if you have questions (212-6126800). . Comments on Securities Lending Facility General Comments: We support the Treasury’s goal of alleviating periods of chronic fails, such as the 2003 episode referenced in Treasury’s white paper, “Consideration Of A Proposed Treasury Securities Lending Facility.” As the Treasury mentions in its white paper, there exists the possibility that the creation of an SLLR “would undermine or reduce private sector incentives to better (and perhaps more efficiently) resolve the issues that the SLLR is intended to address.” We agree that this possibility exists, especially given steps market participants have made in the past year. These steps include: • CBOT position limits instituted • CBOT margin requirements made more stringent • BMA Negative Rate Trading Guideline agreed-upon & released The CBOT’s position limits appear to have reduced the likelihood of supply shortages in the CTD issues. As for negative rate trading, it may be something that could use some time to develop (perhaps with more BMA or Treasury prodding/involvement). If the market gets comfortable processing fails on negative-rate trades, normal market incentives could work to avoid periods of chronic fails. Also, a market-clearing price (& one in which there is significant incentive to make delivery) could also be arrived at via other mechanisms that utilize the existing private market- such as formalizing ‘guaranteed delivery’ (as opposed to 'prompt delivery') trades for which borrowers could ‘pay up’ to obtain greater protection against failing to receive. While ‘guaranteed delivery’ trades may never evolve, it is possible that the other recent changes (CBOT position limits, BMA’s negative-rate trading guidelines) could work sufficiently well, given time and the attention of market participants, to lessen the necessity of an SLLR. Comments On Possible SLLR Structure: If the facility is created, we agree with Treasury’s stated objective that the SLLR be structured such that “it would act as a form of ‘catastrophe’ insurance,” used solely when chronic fails “threaten the overall functioning of the Treasury and broader financial markets.” In addition, we agree that the facility should have minimal impact on the market when it is not being used. Given that the market’s awareness of large supply being available at a certain rate can impact market conditions for an issue (even when the supply is available at rates well below current market-clearing rates), we have some comments on an SLLR’s structure. In our view, these characteristics could help achieve the Treasury’s goals of limiting SLLR’s usage to periods of “extreme market disruption,” and ensuring that the SLLR has limited impact on the market’s normal incentives & operations. • Settlement & term periods: T+5 (or perhaps a longer lag) & a 1-week+ term seem ideal. o The settlement lag could assist in alleviating May 2013-type chronic fails, yet be long enough to limit usage to clearly dire situations.

If a prospective borrower decides not to use the facility due to (for example) T+5 settlement - but would have if settlement had been T+2 or T+3 - then the fail situation likely does not meet Treasury’s definition of “acute and protracted fails.” o Lengthy forward-settlement & term periods would reduce the moral hazard mentioned by Treasury in the white paper, that of excessive risk-taking as investors know they can be ‘bailed-out’ by the SLLR. o The market is more likely to get comfortable with negative-rate repo trading if the SLLR does not set an effective ‘floor’ rate of zero. Longer settlement & term periods increase . the likelihood that dealers will have a chance to become comfortable with the BMA’s recommended negative-rate trading conventions. • Rate: Zero or lower; similar rationale as the comments on above regarding settlement & term periods. • Reporting requirements: If the requirements mentioned in the white paper (“daily cash, repo, and futures positions, and fails to deliver and receive in the security borrowed over an interval bracketing the period of borrowing”) do not already infer this, then we also suggest requiring the same information for the period immediately preceding the borrow request (perhaps for the prior week). This could be beneficial in discouraging active or ill-intentioned use of the SLLR. • Right of substitution: We agree with the suggested bonds vs. bonds lending basis, however offering borrowers the right of substitution may be an unnecessary trait of the transaction. While likely desirable by those who borrow from the facility, the right of substitution may be an accommodation that runs counter to the Treasury’s goal that the facility only being tapped in the most extreme situations.

To wit: If the lack of right-of-substitution is what ultimately stops a dealer from borrowing from the facility on a given day, then it is unlikely that the dealer’s situation is the kind of acute, chronic need for which Treasury intends the facility be used. This material has been prepared by Capital Markets Prime Services and is not a product of Lehman Brothers Research Department. It is for informational purposes only. Lehman Brothers makes no representation that the information contained in this document is accurate or complete.

Opinions expressed herein are subject to change without notice. All levels, prices and spreads are historical and do not represent current market levels, prices or spreads, some or all of which may have changed since the issuance of this document. Under no circumstances should this document be used or considered as an offer to sell or a solicitation of an offer to buy any financial instrument mentioned in it. The products mentioned in this document may not be eligible for sale in some states or countries, or suitable for all types of investors; their value and the income they produce may fluctuate and/or be adversely affected by exchange rates, interest rates or other factors.

Clients are advised to make an independent review regarding the economic benefits and risks of purchasing or selling the financial instruments mentioned in this document and reach their own conclusions regarding the legal, tax, accounting and other aspects of any transaction in the financial instrument in relation to their particular circumstances. Lehman Brothers enters into transactions on an arm's length basis and does not act as advisor or fiduciary to its counter parties except where a law, rule or written agreement expressly provides otherwise. Lehman Brothers and/or its affiliated companies may make a market or deal as principal in the financial instruments mentioned in this document or in related securities, options or other derivative instruments based on them. In addition, Lehman Brothers, its affiliated companies, shareholders, directors, officers and/or employees, may from time to time have long or short positions in the financial instruments, including loans, securities or in options, futures or other derivative instruments based on them.

One or more directors, officers and/or employees of Lehman Brothers or its affiliated companies may be a director of the borrower or issuer mentioned in this document. Lehman Brothers or its predecessors and/or its affiliated companies may have acted as agent or arranger with respect to the loans of the borrowers mentioned in this report, and may have managed or co-managed a public offering of or acted as initial purchaser or placement agent for a private placement of any of the securities of any issuer mentioned in this document within the last three years, or may, from time to time perform investment banking, lending or other services for, or solicit investment banking or other business from any company mentioned in this document. No part of this document may be reproduced in any manner without the written permission of Lehman Brothers. ©2006 Lehman Brothers Inc.

All rights reserved. Member of SIPC. . Joseph Shatz Senior Director, Government/FF&O Strategy Merrill Lynch 4 World Financial Center, North Tower New York, NY 10080 August 9, 2006 Jeff Huther Director, Office of Debt Management Room 2412, Department of the Treasury 1500 Pennsylvania Avenue, NW Washington, DC 20220 Re: Comments on Securities Lending Facility Dear Jeff: We strongly support the introduction of a Treasury securities lending facility and believe that if it is implemented properly, the benefits could be substantial. The purpose of the facility should be to relieve protracted fail situations associated with making delivery into Treasury bond and note futures contracts such as with the 4.875% 2/15/12, catastrophic events such as 9/11 which caused the Treasury to reopen the 5% 8/15/11, and other extraordinary circumstances such as occurred with the 3.625% 5/15/13 as the result of extremely low interest rates leading to a “low cost of fail”. The potential benefits of a well constructed securities lending facility outweigh the potential costs, in our opinion. We understand that there are some negatives associated with its construction. For example, one unintended consequence of the facility could be the possibility of a larger short base in some Treasuries because the shorts would know that they can always borrow the security from the Treasury.

We also realize that there are a number of issues to address from a legislative and regulatory perspective such as changes in Treasury’s borrowing authority, debt limit accounting and tax treatment of borrowed securities. Yet, a properly constructed securities lending facility will protect the efficiency and liquidity of the Treasury market under extraordinary circumstances which could otherwise impair its functioning. This will safeguard investor confidence in the US Treasury markets and thus help fulfill Treasury’s objective of keeping borrowing costs as low as possible over time. Below we present our views on each of the design features discussed in the Treasury Securities Lender of Last Resort (SLLR) Facility White Paper "Consideration of a Proposed Securities Lending Facility." (referred to as “Proposed SLLR paper” .

hereafter).1 In general, we agree with the large majority of the potential design features of the facility suggested by Treasury in their paper. Distribution Mechanism: Auctions versus Fixed-Rate (Price) Standing Facility Securities borrowed from the SLLR, in our opinion, should be fixed in rate with the quantity determined by the borrower. We believe that the Treasury should stand ready to lend an unlimited amount of each issue through the facility. If the amount is limited, then the facility will suffer from the same problem as the Fed SOMA securities lending program; in situations where the market shortage is at an extreme, there will not be enough supply available for lending to prevent protracted fails. Rate, Maturity, Delivery and Reporting Options The facility should be structured in a fashion that allows parties to borrow specific issues versus effectively lending money at an implied repo rate that is relatively close to or at zero. The rate should not be much above zero because the facility should not be designed in a manner that will prevent issues from trading “on special” (i.e.

below general collateral repo). Allowing issues to trade on special is an important natural function of the market determined by supply and demand for individual issues. It is vital that any securities lending facility does not interfere with the normal functioning of the market. In order to ensure that borrowing from the facility will be economical only when delivery fails in a security become chronic or systemic, we recommend a loan maturity of about three to five days (unlike the UK non-discretionary repo facility which is overnight).

The longer (shorter) the loan term, the less (more) the facility will likely be used by market participants. We believe that a three to five day loan maturity strikes a reasonable balance. We recommend same day settlement/delivery to alleviate the liquidity problem as quickly as possible. The Proposed SLLR paper discusses the possibility of a lengthy forward settlement to discourage strategic use of the facility in implementing short-run trading strategies, but we believe that the benefits of a rapid settlement process outweigh the risks. Finally, we agree that requiring borrowers to report daily positions and fails during the borrowing period could be useful in guarding against inappropriate uses of the facility. Collateral We agree that the SLLR should lend securities on a ”bond-for-bond” basis (i.e.

borrower must pledge other Treasury securities of equal market value, plus a margin, as collateral to borrow securities from the facility) in order to avoid affecting Treasury’s cash position. This is the method that the UK DMO non-discretionary repo facility uses successfully. 1 The UK Debt Management Office’s (DMO) non-discretionary repo facility, in our opinion, also provides a relatively good example of how many features of the facility could be structured. . Treasury should allow borrowers to substitute collateral while borrowing from the SLLR. It is important to market participants to be able to substitute collateral in the event that the original collateral begins to trade “on special” (i.e. below general collateral rates) or, for example, if the borrower just needs back the original collateral to make delivery of that specific issue to a counterparty. Inclusion of a tri-party arrangement in the SLLR (as described in the Proposed SLLR paper) would be very helpful. Available Securities The SLLR should stand ready to lend supply of any existing Treasurys, not just specific issues (e.g. on-the-run securities).

It is too difficult to anticipate in advance which issues may have problems which require the use of the facility to address. For example, as the Treasury bond and note futures quarterly contracts expire and new contracts are listed, there are constantly new Treasuries which become the cheapest-to-deliver (CTD). Borrowing Mechanics and Public Transparency We agree that borrowing requests should be submitted to the NY Fed, and that the aggregate daily volume of these requests by CUSIP should be promptly disclosed (with the names of individual borrowers kept confidential). Eligible Borrowers Limiting direct participation in the SLLR to primary dealers, as discussed in the Treasury proposal, makes sense from an operational standpoint for Treasury and the NY Fed. Collateral Margin and Valuation In order to protect Treasury, the amount of collateral (in market value of other Treasury securities) required from the borrower should be equal to the market value of the security borrowed plus some margin. This margin should be calculated as a percentage of the market value of the security borrowed rather than as an absolute dollar amount.

The collateral and the borrowed securities should be marked-to-market daily, and margin calls should be made (or excess collateral returned) as required. Borrowing Limitations There should not be any limit on the amount of securities that can be borrowed by any one participant, in our opinion. As long as there is daily reporting of borrowers’ cash, repo, and futures positions to the Treasury/NY Fed, and also prompt public disclosure of aggregate SLLR positions for each CUSIP (without disclosing positions of individual borrowers), we do not feel there is a need to limit the amount that can be borrowed. Rollovers/Loan Extensions . We agree that it would not make sense to impose significant penalties if borrowers are unable to return borrowed securities to Treasury on the closing leg of the lending transaction. Extending the loan on a day to day basis at an effective implied repo rate of zero percent (i.e. same way private fails are treated) seems appropriate. Sincerely, Joseph Shatz Senior Director, Government/FF&O Strategy Merrill Lynch (212) 449-9196 Joseph_Shatz@ml.com . Wrightson ICAP Harborside Financial Center Plaza V – 12th Floor Jersey City, NJ 07311 Tel: (212) 815-6540 August 10, 2006 Mr. Jeff Huther Director, Office of Debt Management Room 2412 Department of the Treasury 1500 Pennsylvania Avenue Washington, D.C. 20220 Re: Comments on Securities Lending White Paper Dear Mr. Huther: The proposed backstop securities lending facility would be a valuable tool for use in resolving, and indeed forestalling, systemic breakdowns in the clearing process for Treasury securities of the kind seen on several occasions in recent years.

The Treasury delivery process has been smooth in recent quarters, but that does not mean that the risk of logjams has evaporated. We strongly support the Treasury’s efforts to design a low-impact mechanism for ensuring the efficiency of the government securities clearing process. To that end, we have taken the liberty of submitting below a previously published article on this subject in the form of a public comment. Sincerely, Louis vB Crandall Chief Economist LC@Wrightson.com .

Proposed Securities Lending Facility. We’re big believers in the old adage that “if ain’t broke, don’t fix it”. However, we’re also big believers in an often-overlooked corollary to that rule: “If it’s breakable, build a safety net”. The current settlement system for Treasury securities is demonstrably breakable. The severe delivery logjams that developed in 2003 and 2004 haven’t been repeated lately, but the possibility of a recurrence cannot be ruled out.

The market has made tentative progress toward facilitating negative-rate trading1, which might improve the elasticity of the supply of collateral in the financing market during a squeeze, but that doesn’t mean that the threat of future dislocations has been eliminated altogether. There are several possible ways of responding to future squeezes. On the regulatory side, the authorities might bring Treasury buy-in rules into line with those in effect in other, less-favored markets, or they might find a way to force customers to cooperate with dealers in industry-wide trade-netting round-robins in the event of a future logjam. Alternatively, the Treasury might lower the bar for future snap reopenings in response to endemic delivery fails.

All of these solutions seem heavy-handed. A backstop lending facility, in contrast, would provide a temporary safety valve in the event that the clearing system comes under serious pressure again in the future. Unfortunately, the Treasury does not have the luxury of waiting until the need for such a facility once again becomes obvious before moving forward. The process of obtaining the necessary authorizations from Congress and working out the Total Weekly Delivery Fails in Treasury Securities to operational details could easily Primary Dealers take a couple of years.

The Sum of daily failures to receive at dealers Treasury will have to start laying In billions of dollars -- through July 12, 2006 the groundwork now if it wants 1,750 to have such a facility available 1,500 to deal with future cases of settlement gridlock. The 1,250 Treasury will naturally be 1,000 cautious about tinkering with a system that has not caused 750 problems in recent quarters, but 500 we suspect that concerns about 250 the lead-time involved will persuade the Treasury to proceed with the backstop lending 98 99 00 01 02 03 04 05 06 proposal. In responding to the Treasury’s request for comments about how such a program should be structured, the Street has to address one fundamental issue: how far “back” from the market should a backstop facility be erected? Should the program be intended to be tapped, say, once a month? Once a year? Once a decade? Treasury officials have said Backstop Lending Facility Comment Page 2/4 . that one possibility is to set the parameters of the program at levels that would make it uneconomical for dealers to borrow securities in all but the most extreme circumstances. Our own view is that borrowing from the facility should be expensive, but not prohibitively so. As some market participants have pointed out, capping the potential gains on a particular trade may remove an incentive for investors to participate, which ultimately could hurt liquidity. That’s a fair point about financial markets in general, and is relevant to the Treasury market up to a point. For example, that is the reason why the Treasury is scrupulous about not reopening securities opportunistically when markets are functioning well. It wants to leave room for market-makers to smooth out the normal underwriting swings in Treasury yields. At the same time, though, concerns about capping these sorts of arbitrage gains are probably less pertinent to active Treasury issues (on-the-runs, near on-the-runs, cheapest to deliver, etc.) than to almost any other class of security.

Primary dealers report roughly half a trillion dollars of Treasury coupon transactions each day. The vast majority of those trades depend on the fact that individual Treasury securities don’t trade erratically relative to nearby issues. The liquidity of active Treasury securities owes more to their role as hedging instruments and trading vehicles than to the participation of speculative accounts exploiting special situations in individual issues. Placing boundaries on the potential mispricing of individual securities is not the main objective of the Treasury’s proposal, but we don’t view it as a negative side effect.

A lending facility with an effective borrowing rate set, say, at 800 or 1000 basis points below the fed funds target would leave plenty of room for ordinary relative-value arbitrage while protecting market-makers and other participants from the adverse effects of the most severe dislocations. Ultimately, however, the justification for this proposal rests on the need to ensure the integrity of the market’s clearing mechanism. This is important in two respects: Treasury market liquidity. The logjams that developed in 2003 and 2004 forced marketmakers to curtail their exposure, and raised questions about the reliability of the Treasury market settlement process in general among suppliers of collateral to the sec lending market.

In doing so, those disruptions posed a direct threat to the liquidity of the Treasury market. A recent issue of our newsletter2 discusses the role played by the Fed’s securities lending program in allowing dealers to make liquid markets in bills in the face of declining supplies. We think a similar argument would hold for benchmark coupon issues in the event of a new outbreak of delivery failures.

The inelasticity of supply in the issues that are cheapest to deliver into futures contracts has already induced the CBOT to place endof-contract position limits on Treasury futures.3 Those sorts of constraints would proliferate in less formal ways if a shift in market conditions were to produce a new epidemic of delivery fails. Backstop Lending Facility Comment Page 3/4 . Financial market stability. Financial markets have been able to support breathtaking increases in leverage in recent years by making more efficient use of collateral. That is true of asset-backed securities and collateralized derivatives markets in addition to repos. For a given level of capital, the amount of borrowing that can be done on a secured basis in any sector is a large multiple of what can be done on unsecured terms. The stability of today’s secured leverage system depends crucially on the timely, accurate posting of collateral.

Anything that clogs the settlement system can result in collateral mismatches, which represents a more serious systemic danger in the context of today’s balance sheet practices than ever before. Furthermore, current trends suggest that the financial system’s dependence on collateralized financial arrangements will only grow over time. In this context, any obstacle to the smooth exchange of collateral poses unacceptable risks.

The Treasury is unlikely to ignore the risk that the cascading delivery failures of 2003 and 2004 might re-appear, with more damaging results, in the future. Nominal GDP and Total Domestic Credit Market Debt Quarterly observations in trillions of dollars 45 40 Credit Market Debt 35 30 25 20 15 10 GDP 5 0 1970 1975 1980 1985 1990 1995 2000 2005 1 http://www.bondmarkets.com/assets/files/Final%20Negative%20rate%20guideline.pdf 2 http://www.wrightson.com/mmo/2006/07/24/#Bills 3 http://www.cbot.com/cbot/docs/73604.pdf Backstop Lending Facility Comment Page 4/4 . 360 Madison Avenue New York, NY 10017-7111 Telephone 646.637.9200 Fax 646.637.9126 www.bondmarkets.com 1399 New York Avenue, NW Washington, DC 20005-4711 Telephone 202.434.8400 Fax 202.434.8456 St. Michael’s House 1 George Yard London EC3V 9DH England Telephone 44.20.77 43 93 00 Fax 44.20.77 43 93 01 August 9, 2006 Mr. Jeff Huther Director, Office of Debt Management Room 2412 Department of the Treasury 1500 Pennsylvania Avenue Washington, D.C. 20220 Re: Comments on Securities Lending Facility Dear Mr.

Huther: The Bond Market Association (the “Association”)1 appreciates the opportunity to comment on the Department of the Treasury’s (“Treasury”) White Paper “Consideration of a Proposed Treasury Securities Lending Facility” published in April 2006 (the “White Paper”). The Association broadly supports the efforts of the Treasury to address proactively issues that may compromise the high liquidity and efficiency in the market for U.S. Treasury Securities (“Treasury Securities”).

Indeed, we believe, as noted in the White Paper, that the Treasury market is an invaluable national asset.2 However, given the importance of the Treasury market to the nation and financial markets globally, care should be taken to understand fully the costs and benefits of introducing a securities lender of last resort facility (“Lending Facility”) as outlined in the White Paper. A Lending Facility represents a significant structural change in the Treasury market and, while it may add to the tools Treasury already has to address certain market liquidity issues, the ancillary costs, in both increased moral hazard and possible diminution of the specials premium, need to be considered and addressed before moving forward with a Lending Facility. We would also urge Treasury to consider the impact of both the continued study of a Lending Facility and the implementation of such a facility on the market’s continuing efforts to examine and recommend market-based solutions to the liquidity and systemic risks posed by widespread and lasting fails in the Treasury market. We appreciate that Treasury has consistently since August 2005 engaged market participants on whether this change is advisable and we encourage Treasury, as it continues to study this option, to seek further market participant input. 1 The Association represents securities firms and banks that underwrite, distribute and trade in fixed income securities, both domestically and internationally.

Our members are also actively involved in the funding markets for such securities, including the repurchase and securities lending markets. More information about the Association and its members and activities is available on its website, www.bondmarkets.com. The Association is expected to merge with the Securities Industry Association in November 2006. 2 See also Statement of Under Secretary for Domestic Finance Randal K.

Quarles to the Association’s Annual Meeting, May 19, 2006 (“…[a]s we all know, the Treasury market is a critical national asset.”) at http://www.treas.gov/press/releases/js4274.htm. . Mr. Jeff Huther August 9, 2006 Page 2 Executive Summary Given the length of this letter we thought it would be helpful to highlight the Association’s significant observations: â–º The Association supports the objectives and principles outlined in the White Paper for a Lending Facility but believes that it may be more efficient to address them in a way other than through a Lending Facility. The question, put more broadly, is whether there has been a market failure that can best be addressed through a government response, and whether the liquidity of the Treasury market has been negatively impacted by chronic fails. We believe that a governmental policy response is unwarranted because the market, as noted throughout this letter, has made significant progress to address identified market anomalies with respect to price and risk.

Indeed, we believe that there is no evidence that recent incidents of chronic fails have negatively impacted the liquidity of the Treasury market. We believe, as well, that the current regulatory framework for the treatment of aged fails, including increased capital requirements and, in some cases, buy-in requirements provide sufficient additional safeguards to manage the risks posed by chronic fails. â–º The Association believes that while a Lending Facility may offer some marginal protection from the risks posed by continuing widespread fails in the Treasury securities market, on balance, the costs of setting up a Lending Facility and the potential ancillary negative consequences to the normal trading environment outweigh the potential benefits. â–º Changes in market practice and sensitivity to the risk issues raised in recent chronic fails events have lessened the probability that such events will occur with any frequency. â–º The market has adopted practices that, should a large-scale fails event occur, make it less likely that such an event would continue for any significant length of time. The market is also actively exploring and plans to implement market practices that will permit market participants to manage any risk that aged fails pose. â–º In addition, the Treasury has a number of current tools at its disposal that have proved effective during past chronic fails events.

Since we believe that the risk of the occurrence of widespread, chronic and long-lived fails events is remote, current Treasury tools provide further protections against a fails event becoming the catalyst for an unprecedented systemic breakdown in the Treasury securities market. â–º Notwithstanding that we believe, on balance, that the costs of implementing a Lending Facility outweigh the benefits, if the Treasury decides to move forward with a facility care should be taken to address only the risks posed by an extreme market event. In no event should the normal trading environment—an environment that yields the most liquid securities market in the world – be compromised. Indeed, we believe that any Lending Facility should only address catastrophic infrastructure failures that prevent normal settlement. â–º The Association believes that Treasury should pay particular attention to pricing issues.

In particular, to the maximum extent possible the market should be able to take advantage of market-based pricing incentives, including negative rate repo pricing. Any . Mr. Jeff Huther August 9, 2006 Page 3 pricing formula should work on a coordinated basis with current market-based pricing mechanisms. â–º Other terms of the facility should reflect the expectation that the facility would be used only in the most extreme circumstances which market participants expect to continue for a significant period of time and which cannot be mitigated by available market practices. Settlement timing and term of transaction should be such that market participants will find a facility loan attractive only in extreme circumstances. I. Background: Treasury Market Generally The importance of the Treasury market to the global financial system cannot be overstated. There is no fixed-income market that is more crucial to the global economy, nor more liquid, than today’s primary and secondary market for U.S. government securities.3 U.S.

Treasury securities exhibit a high level of liquidity with low transaction costs and no credit risk.4 The liquidity of the Treasury market makes Treasury securities an important hedging vehicle because dealers may sell Treasuries short with confidence of obtaining the shorted security.5 The ability to short Treasuries allows for efficient and transparent price discovery because it allows dealers to manage their positions and thus quote two-way prices. The attractiveness of the Treasury market is global in nature and, as cross-border capital flows increase, the importance of a liquid Treasury market only increases.6 Treasury securities’ importance to individual investors and the federal government should also not be underestimated. Many loan rates are set by reference to Treasury security yields.

As the issuer of the world’s most liquid debt instrument, the Treasury –and thus U.S. taxpayers – benefits from the presence of this liquid secondary market by receiving the lowest financing costs available. As has been extensively noted, economists generally acknowledge that market participants will pay a liquidity premium7 in order to obtain a particularly liquid financial asset.8 The Treasury captures this premium whenever it auctions new securities. 3 For example, in the first half of 2006, daily trading volume as reported by the primary dealers in Treasury securities averaged $ 547 billion.

See Federal Reserve Bank of New York, http://www.ny.frb.org/pihome/statistics/. 4 Robert P. O’Quinn, “Economic Benefits from U.S. Treasury Securities,” Report of the Joint Economic Committee, U.S.

Congress, 2nd Session, Feb. 2002 at 2 (“O’Quinn”); see also White Paper, Introduction. 5 Market participants who short a Treasury security may seek to cover the short in the repurchase (“repo”) and securities lending market. Thus, an active repo and securities lending market in Treasury securities plays an important role in our financial markets and our economy.

The Federal Reserve also uses the repo market in Treasury securities to implement monetary policy. 6 Information on net purchases of U.S. Treasury bonds and notes by foreigners is available at http://www.treas.gov/tic/tressect.txt. 7 See O’Quinn at 2-4; See also Association Comment Letter on Interagency White Paper on Structural Change in the Settlement of Government Securities, available at http://www.bondmarkets.com/legislative/t108001_appenda.pdf. 8 A large and efficient Treasury market allows market participants to develop a “true” credit risk-free yield curve, thereby facilitating more efficient pricing of financial instruments and allowing financial institutions to hedge interest rate risk more effectively. Treasury securities also serve as a good source of collateral and funding for swaps and other derivative transactions.

See O’Quinn. . Mr. Jeff Huther August 9, 2006 Page 4 It is not surprising, therefore, that Treasuries are the most widely held debt securities in the world. It is also not surprising that protecting the liquidity premium associated with Treasury securities is of paramount concern to the Treasury and to market participants. The Association recognizes, therefore, that it is essential that every effort be made to ensure this high efficiency and unique liquidity be preserved during both normal times and during times of market stress.

Market participants have, in the past few years, enhanced operational resiliency to address settlement disruptions, implemented trading practices -- both preemptive and risk mitigating practices -- and worked with industry utilities to limit the risk of widespread and chronic fails in the Treasury market. Work continues to be done and the Association believes that the Lending Facility proposal contributes to the ongoing dialogue between market participants and official policymakers. II. Fails in the Treasury Securities Market A.

Fails in the Normal Course As the White Paper notes, Treasury market safety, liquidity and efficiency may be threatened by elevated levels of fails associated with market dislocation, catastrophic operational disruptions and very low interest rate environments. In order to address the risks posed by elevated and chronic levels of fails, a broader understanding of fails in the normal course is warranted.9 Fails to deliver contracted for securities occur in the Treasury securities market every day. The Federal Reserve Bank of New York (“FRBNY”) began publishing weekly statistics on fails in March 2004. The data includes historical information back to 1990.

This data shows that fails occurred during every week in the period.10 Indeed, fails to deliver involving Treasury securities averaged $11.4 billion per day during the period July 4, 1990 to July 7, 2006. Fails to receive averaged $12.9 billion per day over the same period. A graphic presentation of the FRBNY’s fails data is included in Annex A to this letter. Fails in a normal trading environment occur for a number of reasons.

Operational issues cause the bulk of day-to-day fails. Incorrect trade details result in mismatches, but, generally, these types of fails are resolved very quickly through communications between the counterparties. These fails are usually fully resolved within a day or two of the originally scheduled settlement date. A market participant may also fail to deliver securities that it expected to receive from another counterparty in a different transaction.

This original fail may occur for operational or other reasons, but it can set off a chain of fails that, in some circumstances, returns to the original failing party. Essentially what may happen is that Party A fails to deliver to Party B who then fails to deliver to Party C who had failed to Party A. Given the speed at which transactions occur in the Treasury market, a chain of fails or a round-robin of fails could involve a large number of dealers. 9 This discussion draws on Michael Fleming’s and Kenneth Garbade’s, “Explaining Settlement Fails”, September 2005, available at http://www.newyorkfed.org/research/current_issues/cill-9.html. 10 See http://ww.newyorkfed.org/markets/pridealers_failsdata.html. .

Mr. Jeff Huther August 9, 2006 Page 5 Finally, parties may fail to deliver a security if they have little economic incentive to deliver or to go to the market to borrow or otherwise obtain the security. This can occur as interest rates become very low and the related cost of failing becomes very low. It becomes less expensive to fail than to enter the repo or securities lending market to obtain the security needed for delivery. For these reasons fails never drop to zero and the fact that market participants have developed conventions for resolving these fails quickly and efficiently contributes to the significant liquidity enjoyed by the Treasury market.

Market participants know that slight operational glitches will be resolved quickly. Market participants understand that minor operational issues will not cause them significant and enterprise-wide harm so the incentive to participate actively in the Treasury market remains high. For example, the market has developed a convention allowing a failing seller to make delivery the next business day at an unchanged delivery price.

This convention allows parties to resolve operational glitches quickly without the threat of penalty, thus adding to the attractiveness of this market. Fails data indicate that the gross level of fails even in normal trading environments in the Treasury market has generally risen over the last five years. This is not surprising given the significant growth in both the volume of trading in the cash and repo markets and the dramatic growth in hedge fund community participation in the Treasury market, along with a corresponding growth in prime brokerage clearance, over this time. The attached graph in Annex B plots the settlement fails data against the rise in overall cash market trading volume. Annex D shows the significant rise in repo transaction activity in the recent past. The fact that fails never fall to zero and the ease with which market participants resolve them does not indicate that fails do not impose costs on the failing party.

A failing seller loses the use of the funds it would receive at settlement for the period of the fail. Generally, the seller loses its ability to invest the funds in the general collateral repo market with a return roughly equivalent to the Fed Funds rate. The loss of the time-value of the cash for the period of the fail is usually (except in extraordinary market conditions) sufficient incentive for a failing party to resolve fails quickly.

Thus, any Lending Facility should not seek to eliminate the day-to-day fails that occur in the Treasury market as these fails pose no significant risk to the market or to market participants and, indeed, contribute to the liquidity and efficiency of the market by making the market operationally attractive and efficient. B. Elevated and Chronic Levels of Fails As illustrated by Annexes A and C, the FRBNY’s data on fails also show incidents of significantly elevated levels of fails. Two prominent periods of elevated fails include the period immediately after the 9/11 attacks and during the summer of 2003.

An examination of each of these incidents, and the differing triggering events that led to them, is helpful to understanding when a Lending Facility may be useful and any alternatives to a Lending Facility.11 11 An additional period of elevated fails occurred during May/June 2005. This incident involved the security cheapest-to-deliver into the Chicago Board of Trade’s (“CBOT”) June 2005 futures contract. Because the fails in that instance resulted from a level of futures contracts that exceeded the supply of cheapest-to-deliver Treasury securities, CBOT has imposed position limits on Treasury futures contracts.

See http://www.CBOT.com/CBOT/docs/60840.pdf. . Mr. Jeff Huther August 9, 2006 Page 6 1. Settlement Fails After 9/11 Following the 9/11 terrorist attacks, fails in the Treasury market increased significantly, caused, in large part, by operational and infrastructure problems at significant market participants including inter-dealer brokers and the clearing banks.12 Fails had averaged $1.7 billion a day during the week prior to 9/11 but rose to approximately $190 billion a day during the week of 9/19. This level is in sharp contrast with the previous fails high of $35 billion a day during the week of May 16, 2001.13 On Thursday October 4, 2001, the Treasury announced that it would reopen an additional $6 billion of the on-the-run ten-year note thereby increasing the outstanding supply of the note to $18 billion.

This was the first unscheduled snap auction of a coupon-bearing security since the regularization of note and bond auctions in the late 1970s and early 1980s.14 Treasury clearly had become concerned with the chronically high fails rate and particularly concerned that a continued high-level of fails might begin to affect the performance and efficiency of the market. Under Secretary Peter Fisher stated that “we want to reduce the risk that these settlement problems turn into a much bigger problem to the Treasury market” and that “we wanted to prevent technical problems in the back office from causing wider problems in the pricing of government securities.” The fails data suggests that the reopening had a direct and immediate impact on the level of fails in the market. Daily average fails fell from $142 billion during the week ending October 3 to $63 billion the next week and then to about $18 billion in each of the following weeks. The specials rate for the ten-year note also rose in the immediate aftermath of the reopening.15 While the reopening in 2001 appears to have been effective in lowering the gross level of fails, Under Secretary Fisher was careful to alert the market that reopenings would be extremely rare and that Treasury undertook such extraordinary measures only in response to truly extraordinary conditions in the market.

Well after the situation had been resolved he stated that “never is a long time, so it would be imprudent of me to say that the Treasury will never again hold such an auction. But you should not count on it…We want to rely on [market participants] to reconcile the forces of supply and demand.”16 12 While much of the increase in fails on 9/11 can be attributed to operational problems experienced by significant market participants, including inter-dealer brokers and the clearing banks, there is evidence that some market participants, reacting to the geo-political uncertainty, chose to increase cash balances and withhold Treasury securities from the lending market. This withholding of supply further hampered liquidity.

See “Summary of ‘Lessons Learned’ and Implications for Business Continuity”. Discussion Notes a2 (Feb. 13. 2002), available at http://www.sec.gov/divisions/marketreg/lessonslearned.htm. 13 A full account of fails and settlement problems after 9/11 is available in Garbade’s and Fleming’s, “When the Back Office Moved to the Front-Burner: Settlement Fails in the Treasury Market after 9/11” (“Fleming/Garbade”), available at http://www.newyorkfed.org/research/epr/02v08n2/0211flem/0211flem.html. This discussion of the 9/11 fails incident draws on that account. 14 See Fleming/Garbade. 15 See Fleming/Garbade. 16 Remarks by Under Secretary Fisher before the Association’s Legal and Compliance Conference, January 8, 2002, available at http://www.treas.gov/press/releases/po906.htm. .

Mr. Jeff Huther August 9, 2006 Page 7 After 9/11 all market participants, including the clearing banks and settlement utilities, significantly upgraded backup facilities and operational resiliency. The Association believes that had these operational backups been in place before 9/11, the operationally caused settlement fails would not have occurred in significant size and would not have persisted for an extended period. 2. Fails during the Summer of 2003 Early in the summer of 2003 short-term interest rates reached their lowest level in forty-five years.

By June of 2003 the Fed Funds target rate had been reduced to 1% (and thus the general collateral repo rate hovered around 1%). Interest rate expectations, though, were changing and intermediate-term yields rose sharply. In response, hedging activity, particularly with the on-the-run ten-year note, increased sharply.

Market participants believed that rates would increase. In order to cover short positions, market participants looked to the repo market and the increased demand forced the specials rate on the on-the-run-ten year note down. In addition, the FRBNY had a relatively small position in this note, and thus had a limited supply to lend to the street through its securities lending program.17 Given that the Fed Funds/general collateral rate hovered around 1%, the rate at which special securities traded did not have far to go before the rate hit zero.

As the specials rate approached zero there was little incentive to borrow the note to cover short positions. In fact, those shorting the on-the-run security were buying a low cost option on the expected rise in repo rates. If the specials rate rose they would benefit, but if rates remained static there was no, or very little, cost to continued fails.

As long as the specials rate remained at or very near zero, leveraged market participants had strong incentives to add to short positions, thus adding to the fails in the market. In addition, some market participants with large positions were reluctant to lend into the market because they questioned whether securities could be returned. Thus, less supply – supply that could have alleviated supply/demand imbalances – was available for delivery. The Association has engaged in educational efforts since 2003 –particularly in Asia – to describe for holders of large positions in Treasury securities the dynamics of the Treasury market. We believe that these efforts have encouraged these holders to participate actively in the Treasury lending and repo markets. III. Chronic Settlement Fails and the Market Response The Association shares Treasury’s concern with maintaining the deep liquidity and efficient market-pricing mechanism that characterizes secondary trading in Treasury securities and does not want to see a repeat of the chronic fails situation that developed in 2003.

To that end the Association has worked with its members to address chronic fail issues with respect to risk, pricing and liquidity. Indeed, the Association has focused its efforts on developing preemptive practices that would make it unlikely that a significant fails event would occur. As noted extensively in studies of 17 Information on FRBNY’s portfolio for lending is at http://ww.neyworkfed.org/markets/soma/tnotes.html. . Mr. Jeff Huther August 9, 2006 Page 8 the 2003 incident, our members suffered a number of consequences as fails continued to remain unresolved. Most particularly, regulatory capital requirements imposed lost opportunity costs on dealers as capital requirements increased.18 Firm personnel spent significant time in back-office and industry-wide attempts to resolve fails. In addition, customer relations became strained as fails to receive and deliver in customer transactions increased. The Association agrees with Under Secretary Fisher’s remarks in discussing the snap reopening after the 9/11 attacks: “We want to rely on [market participants] to reconcile the forces of supply and demand.” To this end, the Association has instituted a number of initiatives that we believe will -- collectively, materially and preemptively -- reduce the risk of a chronic fails situation from developing and will, should a chronic fails event develop, provide the market-based solutions, incentives and pricing that will help reduce chronic aged fails and manage the risk posed by such fails.

We have worked closely as well with the Fixed Income Clearing Corporation (“FICC”) on initiatives to reduce fails at the clearing corporation level. The Association believes that these mechanisms address the specific concerns that have led Treasury to explore the Lending Facility. We believe that had the recently developed mechanisms been in place prior to the 2003 event, that event would, most likely, not have occurred to the same extent. A.

Prompt Delivery Repo Trading Practice The Association, through its Funding Division Executive Committee, published a trading practice guideline for “prompt delivery” repo trading in the fall of 2004. The guideline allows market participants to more easily source securities in situations where a particular government security is experiencing high levels of settlement fails. The guideline encourages entities with supply to make a security available by allowing for cancellation of a transaction without penalty if the seller fails to deliver within the prescribed 15 minutes.

Prompt delivery trades now are available to market professionals when market conditions warrant and increase when shortages occur in the market, as was the case with fails associated with the CBOT 10year futures contract in the summer of 2005.19 B. Negative Rate Repo Trading Practice The Funding Division of the Association approved a Negative Rate Repo Trading Practice. Published in March 2006, the trading practice enables the pricing of a repo transaction in a negative rate trading environment, and standardizes practices regarding failure to deliver on a repo transaction where the Repo Rate is negative. The mechanism described in the trading practice consists of resetting the Pricing Rate to the absolute value of the original negative Pricing Rate (i.e.

negative rate 1 is reset to positive rate 1) upon the failure to deliver in the opening leg of a repo. If during the term of a Negative Rate Repo for which the Seller has not delivered the securities and the Buyer has not exercised its right to declare a default, the Seller delivers the securities subject to the transaction, the Pricing Rate will revert to the originally 18 See 17 C.F.R. 240.15c3-1. The prompt delivery trading practice is available at http://bondmarkets.com/assets/files/Revised%20Repo%20Trading%20Practices%20Guidelines.pdf. 19 .

Mr. Jeff Huther August 9, 2006 Page 9 agreed upon Pricing Rate (i.e., the negative rate) but only for the remaining term of the transaction. The Association believes that this trading practice will promote liquidity by allowing market participants to continue to participate in the market during periods when rates for a particular security go negative, and by encouraging market participants who act as a Seller in negative rate repo transactions to deliver the underlying security. The Association has discussed the trading practice with the FICC to ensure its operational capability to flip repo pricing rates.20 We believe that a robust negative rate repo market, with clearly defined practices, addresses one of the significant contributing factors to the elevated fails levels of 2003 and can be used to source and price securities whenever supply and demand imbalances dictate that a significant premium be paid for a particular security.

In addition, the Association’s negative rate repo trading practice when combined with other incentives that raise the costs of continued failing (such as increased capital requirements for aged fails, increased labor costs and fraying customer relationships), should encourage market participants that would view a negative rate trade as uneconomic (and thus continue to fail) to source securities at a negative rate. During the summer of 2003, market participants were reluctant to enter into such trades as the repo buyer, because under conventions in place at the time, repo sellers would have a free option to fail and still get the benefit of negative rate financing. Clarifying the trading conventions will specifically encourage buyers to seek securities in the repo market in a way that didn’t occur in 2003.

Had there been clarity around the trading conventions for negative rate repo transactions prior to 2003, we believe that that fails incident would, most likely, not have occurred to the same extent. C. Fails Margining The Association, through its Government and Federal Agency Securities and Funding Divisions, continues actively to pursue fails margining as a method to reduce the credit exposures to which firms may become subject during prolonged fails events. These credit exposures become acute if a fail event crosses a coupon payment date and the market value of a security changes materially.

Discussions are active in the Executive, Legal and Operations committees of the Divisions regarding the business, legal, operations and systems issues associated with margining fails. Fails margining and the legal infrastructure needed to implement fails margining will be reviewed by the Funding Division Legal Advisory Committee and the Government Division Legal and Compliance Committee in connection with a revision to master trading documentation. The groups are currently considering a trading practice that would recommend margining of fails that have aged beyond a set time limit.

The Association believes that a fails margining trading practice can be available by the end of 2006. D. Buy-in Procedures for Government Securities The Government and Funding Divisions have embarked on an initiative to review and update, as appropriate, the Association’s Buy-in Procedures for Government Securities. Specifically, 20 The negative rate trading practice is available at http://www.bondmarkets.com/assets/files/Final%20Negative%20rate%20guideline.pdf. .

Mr. Jeff Huther August 9, 2006 Page 10 as has been discussed with Treasury staff, the Association continues to investigate whether to update its procedures to provide for a mandatory cash settlement of transactions that have been in a fail status for an extended period of time. The Association believes that a mandatory cash settlement will close-out transactions that remain on dealers’ books, reduce risk, minimize increased capital requirements, and decrease significantly operational labor costs.21 We expect to publish an Exposure Draft of revised buy-in procedures by the end of the third quarter of 2006. E. Operational Resiliency As noted above, subsequent to 9/11 all market participants, including the clearing banks and settlement utilities, significantly upgraded backup facilities and operational resiliency.

The Association believes that these efforts reduce the potential for widespread operational and settlement failures similar to those that occurred after the 9/11 attacks. F. FICC Changes The Association has encouraged FICC to fully implement its fails netting service. The net settlement system of FICC’s Government Securities Division (“GSD”) is being modified to include all daily outstanding failed settlement obligations in GSD’s regular overnight netting process.

After such modification, each participant’s fail obligations for the current business day will be netted with its other government securities transactions having the same CUSIP number and the next business day’s settlement date. A fail will therefore be outstanding for only one day on GSD’s books before entering the net settlement cycle for the next settlement date. Implementation of this service will eliminate clearance obligations by netting outstanding fails versus new activity, thereby lowering clearance costs and reducing participant risk exposure stemming from outstanding fail positions.

FICC’s implementation for the fails netting service is slated for September 2006.22 IV. Lending Facility In August of 2005, Treasury announced at its Quarterly Refunding that it was exploring the possibility of developing a backstop securities lending facility and solicited dealer input. The concept was also discussed at meetings of the Treasury Borrowing Advisory Committee.23 In April 2006, Treasury provided more detail with a straw man proposal for a facility and asked for input on fundamental questions regarding a potential facility.

Specifically Treasury asked for comment on the question of whether establishment by Treasury of a Lending Facility would be an appropriate response to the potential threat of market dislocation, catastrophic operational disruptions and complications associated with historically low interest rates. As 21 The Association previously submitted to the Treasury recommendations with respect to the buy-in rules for Treasury securities. The letter is available at http://www.bondmarkets.com/assets/files/letter_to_treasury_re_fails.pdf. 22 See the FICC Important Notice available here: http://www.ficc.com/gov/notices/GOV159.05.htm?NS-query. 23 The Treasury Borrowing Committee minutes for August 2005 and November 2005 are available at http://www.treas.gov/offices/domestic-finance/debt-management/adv-com/minutes/. .

Mr. Jeff Huther August 9, 2006 Page 11 described in the White Paper, Treasury continues to be concerned that these types of events could lead to distorted prices in the Treasury cash, derivative and collateral markets, and lead to deterioration in dealers’ market-making activities. Treasury is concerned that, if the issues raised by these types of events are unaddressed, the smooth functioning of the Treasury market could be compromised and thus result in increased borrowing costs for the Treasury over time. A. Need for the Lending Facility In order to determine if a Lending Facility is needed, we believe the following issues should be addressed: • What market conditions will a Lending Facility address? • What is the impact of widespread fails on Treasury market liquidity and Treasury’s cost of borrowing? • Will current Treasury tools address supply/demand imbalances? • Will private market solutions efficiently address these issues? In our discussion below we will address each of these issues, but we would first make some general observations about the need for a Lending Facility. The White Paper describes a number of objectives and principles which Treasury believes will need to be incorporated into any lending facility.

The Association supports these objectives and principles but believes that it may be more efficient to address them in a way other than through a Lending Facility. The question, put more broadly, is whether there has been a market failure that can only be addressed through a government response, and whether the liquidity of the Treasury market has been negatively impacted by chronic fails. We believe that a governmental policy response is not warranted because the market, as noted above, has made significant progress to address identified market anomalies with respect to price and risk.

Indeed, we believe that there is no evidence that recent incidents of chronic fails have negatively impacted the liquidity of the Treasury market. We believe, as well, that the current regulatory framework for the treatment of aged fails, including increased capital requirements and, in some cases, buy-in requirements provide sufficient additional safeguards to manage the risks posed by chronic fails. There could, of course, be extreme market conditions not yet identified for which a Lending Facility may be helpful to maintain orderly market conditions where a catastrophic settlement disruption makes it impossible for the market to address imbalances.

The Association believes continued study of those extreme market conditions would be prudent given the importance of this market. In the White Paper, Treasury describes an objective of the Lending Facility to act as a form of “catastrophe” insurance in the Treasury market with minimal impact in normal circumstances. We believe, however, that, notwithstanding attempts to minimize the impact during normal . Mr. Jeff Huther August 9, 2006 Page 12 circumstances, a significant structural change of this magnitude could force material changes in the dynamics of the Treasury market. If Treasury moves forward with the Lending Facility, the greatest challenge will be to create a structure that does not impact the current economics and incentives that the Treasury market enjoys today. Specifically, we believe that a lending facility consistent with the broad parameters described in the White Paper could potentially have a significant adverse impact on secondary market trading, particularly in specials trading in the repo market. Specials trading in the repo market is an important contributor to the overall liquidity enjoyed by the Treasury market.

The specialness of a security contributes to the liquidity premium enjoyed by Treasury at auction. One study has noted that the “size of the price premium at the auction depends on the total number of basis-point days of ‘specialness’ that the security will generate during its life. The security’s total specialness can increase either through the overnight spread increasing or by the security being on special for a longer time.”24 In developing a facility with an implicit floor on repo pricing, Treasury runs the risk of limiting the potential for market participants to profit from positions in on-the-run Treasury securities. This could discourage market participants from taking positions in those securities by reducing the premium associated with the on-the-run securities relative to the off-the-run securities and may limit the amount of securities that might be available in the specials market.

This, in turn, would make shorting the on-the-run securities less attractive as a hedging and risk management tool. During the 2003 chronic fails event it became clear that there was an implicit floor on the specials repo rate of zero. This floor existed notwithstanding that market participants were clearly willing to pay the equivalent of a negative repo rate to the New York Fed’s securities lending program. Now that the Association has developed guidance on negative rate repo trades and that the significant market participants are, on the whole, in a position to implement that trading practice,25 the Association believes that the pricing mechanism in the market is better able to provide efficiently a source for securities than the Lending Facility which, at whatever price it is implemented, will inevitably reprice the floor for specials trading. In addition, and alternatively, any floor for the rate at which specials may trade could, in a very low rate environment, create incentives for market participants to put on short positions merely in order to speculate on the probability that the Lending Facility would be tapped shortly causing a rise in the specials rate.

Market participants will see the short-term opportunity to take advantage of a very inexpensive option and thus increase the volume of potential fails in the system. Behavior during the 2003 incident illustrates the situation. Thus, care should be taken to ensure that a Lending Facility does not provide incentives that would permit market participants to game the system and thus weaken, rather than strengthen, investor confidence in the continued safety, liquidity and efficiency of Treasury markets. 24 Mark Fisher, “Special Repo Rates: An Introduction”, Federal Reserve Bank of Atlanta, Second Quarter 2002, available at http://ww.frbatlanta.org/filelegacy.doc/fisher_2q02.pdf. 25 See the discussion below in Paragraph V.2.(a). .

Mr. Jeff Huther August 9, 2006 Page 13 1. What market conditions will a Lending Facility address? Treasury identifies three types of incidents – severe operational disruption, very low shortterm interest rate environment, and market dislocation -- that may give rise to severe stress in the Treasury market. Simply put, each of these may create supply/demand imbalances that have not been easily remedied in the past.

However, we note that different mechanisms may be appropriate to ease the stresses in different circumstances and that different mechanisms have, in fact, contributed to the easing of fails stresses. While it may be advisable to increase the supply of a security (either through a reopening or through a Lending Facility) in the case of severe operational disruptions, enforcement through market surveillance may be the preferred approach to combat large potentially manipulative positions. In most cases the Association believes that supply/demand imbalances are best addressed through the pricing mechanism in the market.

For example, although the supply-side was addressed immediately after 9/11 by a snap and extraordinary reopening, the regulatory capital charges and the increased ancillary costs imposed by the market, as well as the probable introduction of additional supply because of attractive pricing contributed to the cleanup of the chronic fails in 2003. 2. What is the impact of widespread fails on Treasury market liquidity and Treasury’s cost of borrowing? Before Treasury moves ahead with implementation of a Lending Facility, it should quantify, to the extent possible, the cost to Treasury of widespread and chronic fails. While we agree that both the post-9/11 settlement disruptions and the 2003 chronic fails event imposed costs on market participants through capital charges and labor costs, we have seen little evidence that there was either a cost to Treasury at auction at the time of the events or that a negative perception of the Treasury market lingered after the event. 3.

Will Current Tools Available to Treasury Address Supply/Demand Imbalances? Treasury has a number of tools that have been shown to be effective in addressing the market conditions that concern Treasury. One such tool authorized by the Government Securities Act authorizes Treasury to call for Large Position Reports (“LPR”) to monitor secondary market activity. 26 This authority specifically gives Treasury a potent tool to combat the development of positions in Treasury securities large enough to allow for manipulative behavior that could threaten secondary market liquidity.

LPRs require entities with direct or indirect control of a specified amount in a Treasury security to maintain records in order to facilitate compliance with potential Treasury information requests. LPRs give Treasury a significant window into those firms that are large enough to control positions that could be used in a manipulative manner. Under Secretary Quarles noted his concern that large positions can, but do not necessarily, indicate potential manipulative behavior and intent.27 The Association believes that LPRs, coupled with the surveillance regime instituted after the Salomon incident in the early 90s, are more than sufficient to uncover real manipulative behavior in a changing market 26 See 15 U.S.C.

§ 78o-5. Under Secretary Quarles’ remarks to the Association’s Annual Meeting are available at. http://www.treas.gov/press/releases/js4274.htm. 27 . Mr. Jeff Huther August 9, 2006 Page 14 environment. In addition, LPRs allow Treasury to identify those holders of large positions that have chosen for good economic reasons to withhold supply from the lending market. In addition, Treasury may exercise its authority under the Treasury’s Uniform Offering Circular for the Sale and Issue of Marketable Book-Entry Treasury Bills, Notes and Bonds (the “UOC”) to reopen a security in order to provide additional quantities of securities to the marketplace.28 A reopening may take the form of a standard auction, a “tap” issue whereby the securities are offered to the market on a continuous basis until the shortage is alleviated, or through any other means that the Treasury deems appropriate. As noted above, unscheduled reopenings are extremely uncommon and will be generally undertaken by Treasury to alleviate liquidity concerns only after the normal market mechanisms have failed.29 As noted above, market conditions have caused Treasury to exercise this authority only once since the adoption of the regular and predictable auction schedule. The Association believes that these tools have proven effective in practice and address the risk of manipulative behavior and the risk of severe operational disruption.

Further, while the unpredictability of a reopening may seem to threaten the regular and predictable schedule for issuance that Treasury has fostered,30 we believe that the rarity of such reopening and the extremity of the circumstances would not alter the ongoing benefits of regular issuance nor change market participants’ reliance on Treasury predictability. 4. Will private market solutions more efficiently address these issues? As described more fully above, the Treasury market through the Association has developed and continues to develop market-based solutions to the differing risks posed by chronic fails. We believe that each of these, while not a complete solution in its own right, will significantly reduce the risk of a severe supply/demand imbalance occurring and will reduce both the systemic and entity risk should such severe circumstances occur. We believe that the appropriate economic incentives and market discipline will make necessary supply available and allow demand to source needed securities. B.

Criteria for a Lending Facility We believe, on balance, that the evidence does not support Treasury moving forward with a facility at this time. However, if Treasury determines that a Lending Facility is needed, the facility should meet the following criteria: ● a Lending Facility should only be used in extreme circumstances and only in circumstances of a significant operational settlement disruption; ● a Lending Facility should not reprice or alter the structure of the market; 28 See 31 C.F.R. Part 356. See Under Secretary of the Treasury for Domestic Finance Peter Fisher, Remarks before the Bond Market Association Legal and Compliance Conference (January 8, 2002). 30 See Minutes of the Meeting of the Treasury Borrowing Advisory Committee of The Bond Market Association, May 2, 2006, at http://www.treas.gov/press/releases/54226.htm. 29 .

Mr. Jeff Huther August 9, 2006 Page 15 ● a Lending Facility should not diminish the benchmark liquidity premium of on-therun issues; ● borrowings from a Lending Facility should be at a significantly higher cost than those available in the market; ● a Lending Facility’s terms and conditions and use by market participants should be transparent; and ● a Lending Facility should always be readily available, predictable and nondiscretionary. Again, we believe a few general observations would help frame the discussion. While the Association believes that, in the vast majority of cases, market-based incentives and practices are best placed to prevent elevated levels of fails and to alleviate the risks posed by such fails, a backstop Lending Facility may be worth further consideration for extreme situations such as catastrophic settlement disruptions. This consideration should carefully assess the types of events for which a Lending Facility is intended and receive further input from market participants to ensure that a Lending Facility will be available and will be used only in the most extreme circumstances in which no other less intrusive solution, either from Treasury itself or from market participants, is evident.

Also, we believe that, in developing further the design features of a Lending Facility, Treasury should take into consideration the work that has been done in the market since the recent instances of high levels of fails and ensure that the terms of the facility complement those initiatives. 1. A SLLR should only be used in extreme circumstances and only in circumstances of a significant operational settlement disruption. As noted throughout this letter, this facility should not be a substitute for the natural market mechanisms and its terms should ensure that it is used only during times of extreme stress when other mechanisms, both market-based and those that Treasury currently has, have failed to alleviate significant and unresolved risks to the system. We believe that these circumstances should be limited to significant disruptions to the settlement system and infrastructure. 2.

A Lending Facility should not reprice or alter the structure of the market. As described more fully in Section V.2.a, care should be taken to ensure that the terms of a Lending Facility do not change the current price discovery process in the market or the current structure of this market and the benefits that flow to Treasury and market participants from this structure. 3. A Lending Facility should not diminish the benchmark liquidity premium of onthe-run issues. A structural change of this significance to the market needs to be done with great care. As the liquidity of the Treasury market is unique and the low-cost financing that that liquidity affords .

Mr. Jeff Huther August 9, 2006 Page 16 Treasury is the underlying reason for studying the Lending Facility, Treasury will need to be sure that no ancillary negative effects will be evident in Treasury liquidity. 4. Borrowings from a Lending Facility should be at a significantly higher cost than that available in the market. The Association believes that a Lending Facility should be used only in the most extreme circumstances and only when all avenues to private sector availability of securities have been explored. The Lending Facility should truly be a lender of last resort and should provide incentives, either through its pricing mechanisms or otherwise, for market participants to source bonds in the private market before coming to the Treasury. 5.

A Lending Facility should always be readily available, predictable and nondiscretionary. Given the impact that additional supply of a given security could have on the market and the expectations of market participants, we believe that the availability of securities from a Lending Facility should not be discretionary. Rather, the terms at which securities will be available should be fully transparent and predictable for market participants. In essence, the facility should always be open. 6.

A Lending Facility’s terms and conditions and use by market participants should be transparent. As noted above, in order to avoid potential dislocations in the market, the terms of use and the use of a Lending Facility should be fully transparent. C. Other Potential Costs of a Proposed Lending Facility: Increased Moral Hazard In addition to the above discussion, the White Paper notes a number of benefits and costs that should be considered in implementing a Lending Facility including an increase in moral hazard that may exacerbate highly speculative trading. The moral hazard risk of bailing out and encouraging speculative short positions may be significant and may weaken investor confidence and impact normal trading conditions.

While it may be hard to quantify these costs precisely, Treasury should not move forward until it is clear that the Lending Facility will not create such ancillary problems.31 V. Comments on the Possible Structure and Possible Terms While the Association is skeptical that a Lending Facility can be designed that meets all the objectives and principles outlined in the White Paper, we also believe that continued study of the concept may be warranted to address the limited case of catastrophic operational and 31 The White Paper refers to securities lending facilities that have been implemented in other countries with minimal disruptive effects. We believe that the size of the Treasury market and the uses to which the Treasury market is put make comparisons to other sovereign issuer markets inapposite.

As noted at the beginning of this letter, the Treasury market is global and the significant advantages to holding long positions and hedging through the use of short positions is unmatched in any other sovereign market. . Mr. Jeff Huther August 9, 2006 Page 17 settlement disruptions. Unforeseen events that could serve to cripple the Treasury market and that are not easily addressed by current tools, both official tools and market trading practices, may require a policy response that includes use of a Lending Facility. The discussion below addresses the terms, conditions and other operational details specifically outlined in the White Paper and offers some suggestions that we believe, should Treasury choose to move forward with this proposal, would both address Treasury concerns and minimally affect the market in a normal trading environment. 1.

Auction vs. Fixed-Rate (Price) Standing Facility Should the Treasury move forward with a form of Lending Facility, the Association strongly believes that it should be structured as a fixed-rate standing facility available to all eligible participants with the quantity borrowed determined by the borrower (perhaps with per dealer limits to avoid any one dealer getting an opportunity to dominate an issue). This method is consistent with an approach that would leave to the market a determination as to the appropriate amount that would reconcile supply/demand imbalances.

Further, we believe that any facility that left discretion to the Treasury as to when borrowing would be available could create significant uncertainty in the market and undermine the regular and predictable schedule that has given Treasury securities their benchmark status. However, if Treasury moves forward with a facility that is meant only to be available in the event of a catastrophic settlement disruption, we believe that the criteria for the facility’s availability should be objective and publicly available. 2. Rate, Maturity, Delivery and Reporting Options The Association believes that the pricing structure of a Lending Facility requires careful scrutiny as the pricing structure presents the greatest potential to change the dynamics of the market and to alter the price-discovery mechanisms that contribute to the liquidity of the Treasury market. a.

Implied Rate of Zero Percent As noted above, the Association believes that the Lending Facility, if implemented, should be designed to address only those disruptions caused by catastrophic settlement and infrastructure failures. It should never be available to address market trading conditions as the market pricing mechanism is best suited to address market imbalances. Any pricing structure should reflect the limited scope that we believe a Lending Facility should address. The White Paper suggests that the Lending Facility make securities available at an implied repo rate of zero percent.

The Association believes that a rate of zero percent would institutionalize an effective floor of zero percent on the repo market and would eviscerate the power of a market-based negative rate repo market to reconcile supply and demand through pricing incentives. Setting the rate at an implied rate of zero, without other terms that would increase the cost to a securities borrower (for example, a delayed settlement convention), would destroy the incentive for market participants to source the bonds at a market price that appropriately reflects the securities’ value. We believe, as discussed above, that an implied .

Mr. Jeff Huther August 9, 2006 Page 18 rate of zero raises the serious potential to undermine the specials market in a low-rate environment and that such a pricing structure could make it less attractive to market participants that hold a particular security to make that security available in the specials market. Should the Treasury move forward with a Lending Facility, a pricing methodology that allows market participants to capture the full specials price of a security should be sought. In a low rate environment, there may be little specials value to capture if there is an effective floor of zero percent. There may be little incentive to participate in the specials market in such an environment and there would be no incentive for market participants to continue to develop a robust negative repo rate capability.

As Under Secretary Quarles noted in his remarks at the Association’s recent Annual Meeting, there would be a rationale for reconsidering the pricing structure of the Lending Facility “if market participants took the steps necessary to foster active negative rate repo trading.”32 As described above, the Association and its members have taken significant steps to provide the trading practice and operational infrastructure to allow for robust negative rate repo trading should the pricing environment require it. After the publication of the Association’s Negative Repo Rate Trading Practice, which was endorsed unanimously by the Association’s Funding Division, the Association engaged its primary dealer members to assess the operational readiness of firms to support negative rate repo trading. A majority of this group has represented that they have the operational and systems support in place today to allow for negative rate repo trading.

For the remaining minority that would treat negative rate trading on an exception basis for operational purposes today, it was clear that they were committing the resources to implement the necessary system changes within a short time frame, in most cases within six months from the date of our discussions. The Association is committed to working with our member firms to ensure that systems capability can support negative rate repo trading in a seamless fashion. We believe that this commitment to negative rate repo trading by the primary dealer firms argues strongly against an implied rate of zero percent for the Lending Facility and we believe that the Treasury should clearly state that it will focus on other pricing structures so that firms will complete their operational changes to allow for negative rate trading. b.

Alternative Pricing Structure of a Lending Facility In developing an alternative pricing structure for a Lending Facility, Treasury needs to consider two goals: (1) allow the market pricing forces of the Treasury market to operate unfettered to address supply/demand imbalances as much as possible; and (2) price the securities from the Lending Facility at a level that would only encourage market participants to borrow from the Treasury when there a clear threat to the market’s efficiency. In order to achieve these goals, a Lending Facility should only make securities available at a fee materially higher than the price quoted and available in the market. We believe that such a pricing mechanism would only encourage market participants to borrow from the Lending Facility if securities are, in fact, not clearing due to settlement disruptions. In these market conditions market participants would recognize that Treasury is the only source for securities. 32 See http://www.treas.gov/press/releases/js4274.htm. .

Mr. Jeff Huther August 9, 2006 Page 19 c. Maturity and Settlement Timings The Association further believes that, whatever fee rate is chosen, both the maturity and settlement timings should be set in a way so that market participants will borrow from the Lending Facility only when conditions are extreme and it is expected that they would remain extreme for a significant period of time. For example, delayed settlement of the borrowing -T+5 was suggested in the White Paper – is appropriate and would ensure that a market participant borrowing from the Treasury believes that the fails event is truly chronic and that the ancillary costs of continuing to fail (labor costs, capital costs, and relationship costs) are such that the market participant is willing to wait to secure the securities and that the securities will not become available in the market.

However, there may be circumstances where it may be necessary, because of catastrophic settlement disruptions, to make securities available on an immediate basis. The criteria for making this determination should be objective and publicly available. Furthermore, the logic of severe and extended market disruption would support a structure that requires borrowing for a term rather than for overnight and the Association believes that in order to ensure that the facility is used very rarely, both delayed settlement and term transactions with a term not under a week be utilized. d. Reporting With respect to reporting, the Association believes that if the structure of the facility, particularly the economic, settlement and term parameters, is designed appropriately for use only during extreme circumstances, reporting daily cash, repo, and futures positions, and fails to deliver and receive in the security borrowed over the period bracketing the time of borrowing would be unduly burdensome on dealers that make use of the facility.

As Treasury has noted throughout the White Paper, its intent is that the Lending Facility be used only during extreme market conditions. We believe that during such periods of extreme market stress personnel within firms will be focused on sourcing bonds, maintaining appropriate levels of capital and working with customers. Additional reporting requirements, while a marginal disincentive to using the facility, would not significantly add to the Treasury’s ability to police the Lending Facility’s use.

Treasury continues to have its market surveillance tools, the LPRs and the Interagency Working Group, for example, to guard against possible inappropriate uses of the facility. 3. Collateral The White Paper suggests that the Lending Facility would lend securities on a bond-for-bond basis. The Association believes that this approach serves best the needs of both the Treasury and the market.

As noted in the White Paper, a bond-for-bond transaction would have no effect on the Treasury’s cash position and thus would have minimal impact on Treasury’s cash management and the Federal Reserve’s open market operations. However, should the Treasury adopt a bond-for-bonds collateral approach we would urge that, particularly if the borrowings are for periods beyond one day, a mechanism be put in place that would permit collateral substitution as dealers would not want to tie particular bonds for an extended period. . Mr. Jeff Huther August 9, 2006 Page 20 4. Available Securities Assuming that Treasury designs a pricing mechanism that would attract borrowing from the Lending Facility only in extreme market conditions, we believe that any Lending Facility should not limit availability to the on-the-run sector but should be available to lend additional supply for any outstanding CUSIP. The causes and effects of future extreme conditions are impossible to predict and maximum flexibility would be consistent with the underlying rationale of the Lending Facility to address any severe disruptions in the supply/demand balance in the Treasury market. 5.

Borrowing Mechanics and Public Transparency As borrowings from the Lending Facility would immediately affect the available supply of securities, we believe that full transparency consistent with the information that the FRBNY makes available with respect to its securities borrowing program will be needed. Borrowing from the Treasury would be a crucial piece of market information that would need to be incorporated into prices immediately in order to address the supply/demand imbalance that a Lending Facility would be expected to remedy, and transparency would contribute to the achievement of the Lending Facility’s objectives quickly. In addition, transparency would prevent, as the White Paper noted, the borrowing firm from gaining an informational advantage over other market participants. 6.

Eligible Borrowers We agree that any Lending Facility should be limited to primary dealers. The primary dealers are the significant market makers in Treasury securities and have the necessary experience in dealing with the FRBNY on these types of operations. Full transparency of the rate at which securities will be available will allow other market participants to instruct a dealer to borrow on its behalf. 7.

Collateral Margin and Valuation The Association agrees that appropriate margin, at levels consistent with market norms, should be maintained at the Treasury, as this reflects the market convention for these types of transactions. 8. Borrowing Limitations The Association believes that specific borrowing limitations should not be necessary if the facility is designed in a way that encourages its use only in extreme market conditions. By its terms, the Lending Facility should be uneconomic for purposes other than remedying a large scale disruption.

Treasury could continue to use its other market surveillance tools to ensure that any large volume borrowings are for a purpose reflecting the goals of the Lending Facility. . Mr. Jeff Huther August 9, 2006 Page 21 9. Rollovers/Loan Extensions Again, in keeping with the Association position that any Lending Facility should mirror, to the extent consistent with its goals, standard market practices for repos and securities borrowings, we believe that fails back to the Treasury should not be treated punitively. Creating unequal incentives for failing to different classes of counterparties could exacerbate a fails situation. Thank you for the opportunity to respond to the Treasury’s proposals as outlined in the White Paper. The Association and its membership looks forward to working with Treasury on the issues raised in the White Paper.

Please feel free to contact Robert Toomey (646.637.9224 or rtoomey@bondmarkets.com) at the Association should you have any questions or comments regarding our response. Sincerely, /s/ Stephen G. Malekian /s/ John A. Roberts Stephen G.

Malekian, Managing Director Citigroup Chairman Funding Division Executive Committee John A. Roberts, Managing Director Barclays Capital Inc. Chairman Government Division Executive Committee /s/ Robert B. Toomey Robert B.

Toomey, Vice President and Assistant General Counsel The Bond Market Association Principal Staff Advisor to the Funding and Government Divisions cc: Treasury Department: Randall Quarles, Under Secretary for Domestic Finance Emil Henry, Assistant Secretary for Financial Institutions Bureau of Public Debt: Lori Santamorena, Executive Director Michael Sunner, Deputy Assistant Counsel Federal Reserve Bank of New York: Dino Kos, Executive Vice President Debbie Perelmuter, Senior Vice President Joyce Hansen, Deputy General Counsel and Senior Vice President Michael Nelson, Vice President The Bond Market Association Funding Division Executive Committee Government Division Executive Committee Primary Dealers Committee Micah Green, President Randy Snook, Executive Vice President John Vogt, Executive Vice President . August 11, 2006 BY FEDERAL EXPRESS AND E-MAIL Mr. Jeff Huther, Director Office of Debt Management Room 2412 Department of the Treasury 1500 Pennsylvania Avenue, NW Washington, D.C. 20220 Re: Comments on Securities Lending Facility – 71 Fed. Reg.

26174 (May 3, 2006) Dear Mr. Huther: The Board of Trade of the City of Chicago, Inc. (“CBOT®” or “Exchange”) appreciates the opportunity to comment on the Department of the Treasury’s (“Treasury”) request for comment regarding whether it should establish a securities lender of last resort facility (“SLLR”). The CBOT is one of the world’s leading and most liquid derivatives exchanges based on contract volume.

Our flagship U.S. Treasury products include futures and options on 30Year U.S Treasury Bonds, and futures and options on Ten-Year, Five-Year, and TwoYear U.S. Treasury Notes.

The Exchange’s U.S. Treasury futures and options play an important role in the interest rate risk management strategies of a number of foreign and domestic market participants, including governmental entities, banks, insurance companies, pension funds, and mutual fund managers, among others. Approximately 535 million futures and options contracts on U.S.

Treasury securities were traded on the CBOT in 2005, representing a notional value of approximately $55.6 trillion. The Exchange shares the Treasury’s interest in the maintenance of a safe, liquid and efficient U.S. Treasury securities market. At the same time, the CBOT desires to continue to be able to provide fair, efficient and orderly U.S.

Treasury futures and options markets. The Exchange commends the Treasury’s continued proactive approach to promoting market efficiency in the Treasury securities market and the broader financial markets through its consideration of a proposed SLLR. The liquidity of the cash Treasury market and the ease of arbitrage between the cash and futures markets can either facilitate or impair the process of convergence of cash and futures prices. It is precisely this convergence of the two prices that makes a futures contract relevant and useful for market participants.

The increased incidence of fails in recent years in the cash Treasury market, and especially in the Treasury repurchase market, has eroded trade certainty in the cash Treasury market in ways that impair the . Mr. Jeff Huther August 11, 2006 Page 2 of 2 ability of market participants to perform cash-futures arbitrage. Therefore, we believe that the Treasury’s implementation of an SLLR would have a positive impact on the Treasury securities and futures markets by making available an additional, temporary supply of Treasury securities in those instances when market shortages might threaten to impair the functioning of these markets. The CBOT stands ready, in its role as the leading U.S. Treasury futures and options marketplace, to assist the Treasury with respect to the appropriate structure, including terms and conditions and other operational details, at such time as the Treasury determines to move forward in implementing an SLLR. If you have any questions, please feel free to contact Anne Polaski, Assistant General Counsel, at (312) 435-3757 or apolaski@cbot.com. Sincerely, Bernard W.

Dan . . . . . .