April 19, 2006
JS-4191
Remarks of James Clouse,
Acting Deputy Assistant Secretary for Federal Finance
U.S. Department of the Treasury
Before the Information Management Network PAPERS
Forum
Harrisburg, Penn.- Thank you. I'm delighted to have the opportunity to speak with you today. As Al mentioned, I am currently serving as the acting deputy assistant secretary for federal finance while on leave from my permanent position at the Federal Reserve. Much of my work at the Treasury focuses broadly on Treasury debt management issues, so with your indulgence, I would like to devote my time this morning to conveying a sense of the key policy and operational issues in this important area.
Treasury Market Overview: For the purpose of my remarks today, I will define Treasury debt management as encompassing the entire range of policy and operational issues associated with financing U.S. government operations taking as given the paths for expenditures and receipts. This definition thus excludes all the complex and thorny economic and political judgments involved in tax and spending policies, but includes the panoply of issues related to the range of securities offered; the tactics, strategy and communications around debt issuance; and matters related to the efficiency, liquidity, and robustness of the primary and secondary Treasury markets. Taking this definition on board, effective Treasury debt management is certainly a critical function for the U.S. government, the U.S. economy, and the U.S. taxpayer--though admittedly it is not one that often occupies the front pages.
Some basic statistics testify to the importance of Treasury debt management. As of December 2005, marketable U.S. Treasury debt held by the public totaled about $4 trillion or 30 percent of nominal GDP. Short-term Treasury bills comprise about a quarter of all marketable Treasury debt outstanding. On average every week, the Treasury rolls over roughly $50 billion in maturing bills. Another 65 percent or so of outstanding securities is in the form of nominal coupon securities and about 8 percent is in the form of inflation-indexed securities or TIPS. As everyone in this audience well knows, secondary market trading volume in Treasury securities is enormous. The Fixed Income Clearing Corporation, which compares and nets most trades in the Treasury market, reports daily trade volumes settled of roughly $500 billion. Daily trading volumes in the corporate bond market and equity market are much smaller by comparison. The extensive secondary market trading activity in Treasury securities is accompanied by robust trading in repo markets and in exchange-traded and over-the-counter Treasury derivatives markets as well. In short, the Treasury market broadly-writ is a cornerstone for global fixed-income markets. Treasury securities are highly valued as assets free of credit risk, and the liquidity of Treasury cash and derivatives markets has made Treasury securities the instrument of choice for many investors in managing interest rate risk. The Treasury market thus affords the U.S. government unparalleled access to global financial markets at the lowest possible cost over time. Everything we do as debt managers is focused on financing government operations in ways that also foster the continued growth and efficiency of Treasury markets.
The Financing Environment: The environment in which Treasury debt managers operate is highly uncertain. And the focal point of that uncertainty is the balance of the sources and uses of funds for the U.S. government. At the most basic level--leaving aside some complexities of government accounting--the principal sources of funds raised each day by the federal government including tax receipts and new debt issues must cover the various uses of funds including government expenditures and principal and interest payments on outstanding securities. Any shortfall in the sources of funds relative to the uses of funds must be met by a drawdown in cash assets. Conversely, any excess in the sources of funds relative to the uses of funds implies an increase in cash assets. The Treasury maintains cash assets in several forms including non-interest earning balances held at the Federal Reserve and collateralized deposits at commercial banks; it is generally to the Treasury's advantage to keep cash balances at low levels while at the same time avoiding overdrafts at commercial banks or the Federal Reserve.
The sources and uses constraint is inescapable as are the attendant uncertainties, some of which are most evident at high frequencies and while others are more apparent over the longer-run. At high frequencies, debt managers face great uncertainty about receipts of all sorts but particularly about business and individual income tax payments. Of course, today is a perfect example of that high frequency uncertainty we have some information about this year's tax season, but we really won't have a firm handle on the April/May tax receipts picture until the end of this month. Over the longer run, uncertainties about economic prospects and changes in the fiscal outlook are something that debt managers must constantly confront. The Administration, Congress, and a range of private sector forecasters produce regular deficit projections. But these forecasts can change quite significantly over time, even for the current year.
Goals and Objectives: At first blush, this sort of decision-making under uncertainty doesn't seem to pose any special problems. A textbook analysis might simply say that debt managers should maximize the expected value of some objective function given the sources and uses of funds constraint and the uncertainty about the key fiscal variables. But what objective function should debt managers maximize? The Treasury does not operate with an explicit statutory debt management objective. And it may not surprise you to learn that economists studying this problem have arrived at quite different conclusions over time about the appropriate objectives for debt management. For example, in so-called Ricardian-equivalence frameworks, households view any increase in the value of their existing holdings of federal debt as perfectly offset by a rise in future tax obligations. In such models, choices about the composition of federal debt are largely irrelevant and by implication, debt managers are mostly irrelevant too. For obvious reasons, my colleagues and I are not particularly fond of these models! Other models have suggested that debt managers should concern themselves with structuring debt so as to minimize the variability in tax rates. Still others suggest that Treasury debt should be structured so as to assist in stabilizing the macroeconomy.
As a practical matter, debt managers in the United States and in many other countries have adopted a more pragmatic approach. Broad macro objectives such as stabilizing tax rates or the economy are viewed as the province of Congress and the Administration. The pragmatic approach focuses largely on minimizing the cost of debt financing over time while also giving some weight to other factors such as variability in interest payments, the diversification of the investor base, operational risks, and flexibility. To some extent, the pragmatic approach is amenable to standard analysis of decision-making under uncertainty. For example, given the presence of a term premium in longer-term Treasury yields, a strict cost minimization criterion would probably favor--at the margin--shorter-term financing. On the other hand, yields on longer-term securities are less variable over time than short-term yields and longer-term securities can be employed to lessen rollover risks. An optimal debt portfolio might then involve some appropriate balancing of these considerations.
While a useful exercise, there is a fundamental difference between this standard sort of analysis and Treasury debt management in practice: The textbook exercise assumes that Treasury is a price-taker in financial markets, but in reality the strategies and tactics employed by Treasury debt managers can affect the market prices of Treasury securities. It is this interplay of market expectations and debt management policy that makes Treasury debt management more subtle and nuanced than a simple optimal cash management exercise. As I will discuss more in a moment, transparency and adherence to regular and predictable issuance polices are the principal components of the Treasury's strategy in managing the linkage between market expectations and debt management policy.
Regular and Predictable Issuance: As a general principle, longer-term debt is issued primarily to address longer-term financing needs while short-term debt is issued to address transitory financing requirements. In practice, this means that issuance of longer-term securities tends to be smooth over short horizons. Such "regular and predictable" issuance for longer-term securities allows investors to better anticipate prospective supply over the near-term. That makes good sense because the demand curve for longer-term securities is often assumed to be more steeply downward sloping than that for short-term securities. As a result, the reduction in uncertainty about supply afforded by regular and predictable issuance of longer-term securities results in lower long-term interest rate volatility and a lower interest rate risk premium for long-term Treasury securities. The Treasury's sources and uses constraint, however, implies that not all debt issuance can be regular and predictable--something must act as the shock absorber in contending with irregular and unpredictable expenditures and receipts. Treasury bills (including cash management bills) tend to fulfill this shock absorber role, responding to seasonal and one-off funding needs. This again makes good sense because the demand curves for shorter-term debt are viewed as fairly flat. As a result, the uncertainty about the supply of bills has only a modest impact on the volatility of short-term yields and the risk premium that Treasury must pay on short-term securities.
These observations bear on an important topic that we face on an ongoing basis. One of the most common types of comments that we receive from market participants is "why don't you issue X" where "X" might be very long-term debt, or floating-rate debt, or debt with certain embedded options, or foreign-currency denominated debt…the list goes on and on. Usually these suggestions arrive at times when the market pricing for security X is particularly rich, so that it may appear that the Treasury is forgoing an opportunity to lock-in inexpensive funding. However, this sort of reasoning overlooks several key points. First, the Treasury aims to issue securities in a way that fosters liquidity in active secondary markets. Active market trading, in turn, leads to strong auction demand for securities and lowers Treasury's borrowing costs over time. As discussed above, these objectives generally lead to stable and predictable issuance patterns over time. In effect, the decision to issue a particular type of security is much like a capital investment decision--one must consider the demand for that security over a long horizon, not just at the moment. Second, opportunistic issuance of certain special securities to meet a transitory demand could well entail significant costs. Funds raised in this way would, through the Treasury's sources and uses constraint, imply a cutback in funds raised from other sources. The resulting increase in uncertainty in funds raised from those sources that were trimmed could prove detrimental to Treasury's overall funding cost. In addition, a policy of regular opportunistic issuance would no doubt prompt a great deal of market speculation about whether the Treasury would or would not be entering various markets. The resulting uncertainty and volatility could hamper the development of these market segments over time.
Squaring Supply and Demand: So far, I've portrayed Treasury issuance strategies as largely supply driven--that is determined by the federal government's financing needs. Obviously, that's only half the story with the Treasury's assessment of the demand side of the market naturally being the other half. Assessing market demands is inherently difficult, and we rely on the counsel of market participants and conversations with groups like this one in reaching those judgments. Prior to each quarterly refunding auction, we visit with a number of primary dealers to gain their perspective on current market conditions. In addition, we regularly seek the advice of the members of the Treasury Borrowing Advisory Committee (TBAC), which includes senior representatives from a range of financial firms, on market conditions and the demand for Treasury securities across different market segments.
We also rely on our visits with groups like this one that have special knowledge about demands in particular market segments. Indeed, one of the reasons I wanted to address this group today is to make an unabashed pitch for Treasury securities as you contemplate your asset management decisions. According to the Federal Reserve's Flow of Funds Accounts, pension funds as a group have not been large holders of Treasury securities; private- and state and local government pension plans hold only about 2 percent and 4 percent, respectively, of their total assets in Treasury securities. With the reintroduction of the thirty-year bond, and a full slate of regular inflation-indexed securities, we feel that our range of offerings should be attractive to pension funds and we hope to see those aggregate percentages edge up over time. We would also welcome greater participation by pension funds at Treasury auctions as part of that process.
Conducting Auctions: Up to this point, I have focused largely on basic conceptual issues regarding supply and demand conditions, but ultimately debt managers must go about the practicalities of selling their wares. As part of this process, Treasury staff carefully review the incoming data on hundreds of expenditure and receipt categories on a daily basis and update their near-term projections of financing needs accordingly. Based on these projections, Treasury debt managers then sketch out a path for issuance that meets the projected financing gap. At weekly meetings, Treasury staff present recommendations for the quantities to be auctioned during the subsequent week. The amounts offered are announced on the public website of the Bureau of the Public Debt at about 11 a.m. Immediately thereafter, trading in the so-called when-issued market begins in earnest. The Treasury's Bureau of Public Debt (BPD) and the Federal Reserve Bank of New York collect both competitive and non-competitive bids. On the day of the auction, BPD and Fed New York staff accept competitive bids until the close of competitive bidding at 1:00 p.m., with a large fraction of the total volume of bids for any auction submitted in the last 15 minutes of the bidding period. After the auction close, the BPD publishes auction results on its website including the auction stop out rate, bid-cover ratio, and bids and awards for primary dealers, indirect bidders, and direct bidders. As part of our ongoing commitment to efficiency in the auction process, the Treasury has reduced the time it takes to release auction results to 2 minutes or less following the auction close.
It is worth noting that the resilience of the Treasury auction system has been dramatically improved in the years following the 9/11 attacks. Moreover, to further modernize the auction system, the Treasury is currently in the initial stages of designing a new auction system that will continue to enhance the efficiency and flexibility of Treasury auctions.
The Secondary Market: The job of Treasury debt managers does not end once a security has been auctioned. Debt managers play an active role along with colleagues in other agencies in monitoring developments in cash, repo, and Treasury derivatives markets. We are particularly attuned to factors that could threaten the efficient functioning of the Treasury market and thereby undermine the special role of Treasury securities in fixed-income markets. For example, at times in the past few years, the Treasury market has been beset by very large volumes of delivery fails. Delivery fails occur on a regular basis in the Treasury and other financial markets, but in the period after 9/11 and in the low interest rate environment of 2003, delivery fails spiked to very high levels. Fails at these elevated levels threaten the efficiency of the overall Treasury market and, at times of stress, may exacerbate a sense of panic in the market. For these reasons, the Treasury has been studying for some time the possibility of establishing a securities lending facility. There are many pros and cons associated with such a facility and the Treasury has not taken a position on whether a securities lending facility is appropriate public policy. In an effort to facilitate public comment, the Treasury plans to publish a discussion paper on this topic.
Another matter that officials must bear in mind when reviewing secondary market conditions is the potential for a single firm to gain effective control over the supply of an outstanding issue. Consolidation over recent years has implied that an increasing number of financial firms have the wherewithal to assume very large positions that may end up conferring effective control over particular Treasury securities. The difficulty here comes in distinguishing these cases from those in which scarcity value for a particular security arises from more benign market forces. As always, it is important for senior management of all firms that are active in the Treasury market along with their firm's compliance officers and risk managers to ensure that any especially large positions in Treasury securities are consistent with all applicable rules and regulations.
Conclusion: To conclude, I hope that I have convinced you of the proposition advanced at the outset of my remarks this morning--that effective Treasury debt management is a critical function for the U.S. government, the U.S. economy, and the U.S. taxpayer. It is also one that presents a great many challenges, both intellectual and operational. We are committed to meeting these challenges, and have been hard at work on many fronts to foster an even deeper, more liquid, and more efficient Treasury market. The insights obtained in visiting with groups such as this one are an integral part of that overall process. Thank you very much.