FROM THE OFFICE OF PUBLIC AFFAIRS

The resilience of the U.S. economy has surprised even its admirers.

One source of this resilience is the federal government's role as a shock absorber. When economic activity slows, taxes fall and expenditures rise, both automatically and as a result of targeted legislation. From physics we know, however, that every action has an equal and opposite reaction. So while for a given sector of the economy, or set of incomes, the federal government can absorb a shock, this only happens by transferring it somewhere else. The ultimate cost is passed to the broader risk pool of federal taxpayers. The immediate cost - in terms of funding - is transmitted from the real economy back to the financial markets as variance in the federal government's borrowing requirements.

You contribute to the economy's resilience as well. You price the risk of likely and unlikely outcomes for the value of commodities and products and companies and help to transfer these risks to those most willing to absorb them. You also price the risk of likely and unlikely outcomes for the federal government's borrowing needs.

It is one of my jobs to manage the fluctuations in the government's borrowing needs. It is one of your jobs to price that risk. So I thought I would take this opportunity to explain how I see the objectives and the constraints of the Treasury's debt management in the hope that it might be of some use to you.

By way of illustration, I will say a few words about our decision last fall to suspend issuance of the long bond.

But, first, let's talk about the economy.

Both Glenn Hubbard, the chairman of the President's Council of Economic Advisors, and Alan Greenspan, have said that an economic expansion is underway. GDP growth in the fourth quarter of 2001 was recently revised up to 1.4 percent, so that it now appears that there was only one quarter of contraction in 2001 - the third quarter, which included September 11th. Of course, even revisions have a way of being revised themselves, but signs point to improvement.

Of particular significance, I think, is the fact that estimates of productivity for the fourth quarter were revised up again to an astounding 5.2 percent annual rate. This is almost surely above that possible in the long run but it may indicate that the long-run growth rate is higher than we previously thought.

The outlook for 2002 thus looks fairly bright. In the first half of this year, we expect the economy to benefit from a swing in the inventory cycle and a continued growth in household and government spending. In the second half of the year it will be crucial that business investment spending rebound from the path of contraction that has been evident since the summer of 2000. There are reasons to be optimistic. In recent months shipments of non-defense capital goods have been increasing. Also, the passage of the stimulus bill that was just signed into law by President Bush will provide significant new incentives for corporate investment.

Although most of the indicators are favorable and recovery appears to be under way, we must not forget that more than a million Americans lost their jobs since the recession began. As Secretary O'Neill stated recently, the stimulus bill "will add momentum so that we have a more robust recovery and return to full prosperity. [The bill] will speed Americans back to work and help the unemployed until they return to work."

Both the extension of unemployment insurance and the new tax incentives for corporate investment will have to be absorbed, in the first instance, in our borrowing requirements. We know that in times of war, national emergency or recession it makes economic sense for the federal government to run deficits. We also know that we would like to return the federal government to a surplus position and that this will happen as our economy gathers momentum and discipline is exerted in the budget process.

The recent swing from large surpluses to deficits and our objective of returning to surpluses, once again, all serve to cast a spot light on the role of debt management and the impact that variance in the federal government's borrowing needs has on our financial markets.

For the last quarter of a century, the Treasury's debt management has been described as serving three objectives: first, the lowest borrowing cost over time; second, efficient management of cash balances; and, third, the promotion of efficient capital markets.

For many years, these three stated objectives appeared to be complementary and market participants have often thought of them as self-reinforcing. In practice, however, debt management has always involved trade-offs among these three objectives and these trade-offs have become more evident in recent years.

From the late 1970s to the mid-1990s, we faced seemingly ever-expanding deficits and borrowing needs. During this period, Treasury developed its pattern of regular and predictable issuance of a wide range of securities in order to promote efficient capital markets and, by doing so, also serve our objective of borrowing at the lowest cost over time. But, given the lumpiness of federal tax receipts, regular and predictable issuance has always posed a challenge for efficient cash management. For the most part, variance in receipts has been absorbed by expanding and contracting cash balances. Only at the margin have we deviated from regular and predictable issuance - through seasonal changes in bill issuance and the use of cash management bills - to limit the swings in cash balances.

In the late 1990s, as surpluses rapidly materialized and it became necessary to reduce issuance, the tension became obvious between promoting capital markets, on the one hand, and achieving the lowest cost borrowing, on the other. Market participants habituated to the use of Treasury securities for pricing and hedging and for their own cash management perceived continuation of established issuance patterns as desirable for the promotion of efficient capital markets. But sustaining those issuance patterns, in sizes necessary to maintain the liquidity of each maturity, would have added unnecessary borrowing costs and without a large buyback program would have converted our cash management into an asset management function.

I am dissatisfied with the conventional trilogy of objectives because - as expressed - they give no guidance as to how we will make trade-offs and choices among them and, thus, I fear they will not serve us well in communicating how debt management is conducted in the current uncertain environment. Both as a description of past debt management actions and as a guide for understanding our future behavior, I believe that the same elements can be restated to express the idea that the Treasury's debt management serves a single, overriding objective and confronts multiple constraints.

Simply put, the objective is to meet the financing needs of the federal government at the lowest cost over time.

The dominant constraint that we confront in achieving this objective is that we see the future only imperfectly. We are always making decisions in conditions of uncertainty.

As a consequence, debt management necessarily involves three judgments: first, about what will be the likely size and duration of our borrowing needs, second, about how we should respond if actual needs differ substantially from expectations and, third, about what will be the lowest cost means of financing those needs in the future. We cannot escape these three issues. We face them in our weekly financing decisions, in our quarterly refundings, and in our strategic planning.

Cash management is better thought of as a constraint on our actions, not an independent objective. We need to pay the government's bills as they come due even though our cash flow from tax receipts varies considerably from week to week. We must have enough cash on hand to meet the expected and unexpected variance in both revenues and expenditures even when doing so imposes added costs. But we seek to minimize the extent to which our objective of the lowest borrowing costs is burdened by the constraint of our cash management.

The promotion of efficient capital markets is important, but should not be thought of as an independent objective of Treasury's debt management. In the long run, we know that we need efficient capital markets in order to sustain our ability to finance the federal government. Efficient capital markets are a means to the end of lowest cost borrowing over time. In the short run, however, where we all live, the need to promote and sustain efficient capital markets can act as a constraint on our objective of the lowest cost borrowing.

For example, the unscheduled reopening of the 10-year note last October was undertaken because of concerns about the long-term consequences of systemic failure in our credit markets - even though the uncertainty it engendered may have added to our borrowing costs in the short run. For that reason, unscheduled reopenings will remain the exception - the exceedingly rare exception.

Similarly, the Treasury's continuing commitment to a schedule of regular and predictable auction dates is a means, over time, to the end of the lowest cost borrowing. In the short run, however, this commitment serves as a constraint: with regular and predictable auction dates we accept the cost of occasionally borrowing when it is inconvenient or expensive in return for the lower costs, over time, from providing greater certainty to the Treasury market.

There are other constraints. For example, the availability of the full faith and credit of the United States as a savings vehicle should not be limited only to those who can afford the minimum one thousand-dollar denominations available in our auctions of marketable securities. Thus, we will continue to offer savings bonds even though they are not the most efficient form of borrowing in operational terms. But, again, we will seek to minimize the cost of this constraint as it weighs on our objective by striving for more efficiency.

The framework I am describing, of a single overriding objective and of multiple constraints, informs the decisions of the debt manager. There is an entirely different discussion about positive externalities and the optimal level of government debt for the purpose of financing the federal government, for the functioning of our financial markets and of our economy. For my part, I doubt that zero is the right number for federal debt outstanding. But that policy debate needs to take place away from the explanation of the debt manager's reaction function, away from the effort to explain how we manage the variance in the federal government's borrowing needs as we receive them, day by day.

In explaining this process, I want to underscore the importance of the three judgments we are always making. To achieve our objective of the lowest borrowing costs, we want to maintain a pattern of regular and predictable issuance of as broad a portfolio of instruments as is consistent with (a) our best projections of likely borrowing requirements and (b) our ability to respond if those projections are not realized, and (c) our current understanding of what will provide the lowest borrowing cost over time.

By way of example, you know that in my judgment continued issuance of the long bond was not consistent with our objective, nor compelled by our constraints.

Last October, given the likely path of our borrowing needs over the coming decade, we could not sustain continued issuance of our complete portfolio of instruments. Because we want to maintain the liquidity and depth of the instruments we issue - as a means of achieving the lowest borrowing costs over time - we suspended issuance of the 30-year bond so we could concentrate our borrowing needs on our other instruments. Consolidating our long-term borrowing at the 10-year point is the most effective way for us to maintain a reasonable yield curve and to provide the supply necessary for adequate liquidity.

At that time, it seemed to some as if the economic downturn would be extended and the recovery would be slow. Now it seems that these views may have been too pessimistic. In making our decision to suspend the 30-year, we were neither optimistic nor pessimistic. We simply made a judgment about the most likely path of our borrowing needs.

In addition, we considered the likely consequences of the unlikely outcomes - that is, the situation we would find ourselves in if our projections were not realized and how we would respond. On one side, we faced the risk that we return to surpluses even more quickly than we expected. In this event, maintaining issuance of the 30-year would impair our ability to maintain a portfolio of liquid instruments and prove unnecessarily costly to the taxpayer.

On the other side, we faced the risk that sustained surpluses would not materialize as promptly as we expect. As I explained last October, "if later in this decade it turns out that 30-year borrowing is necessary to meet the government's financing needs, it is still likely that our decision to suspend 30-year borrowing at this time will have saved the taxpayers money. In addition, the reintroduction of the 30-year bond, at some in the future, if necessary, would likely be costless to the Treasury."

I don't expect that to happen - and you shouldn't either - because we also made a judgment about the cost effectiveness of the long-bond, over time. For over a decade, market participants have been telling the Treasury that demand for the bond was insufficient to achieve our objective of the lowest cost financing. Investors simply wanted too high a premium for this added cost to be a sensible means of minimizing our refinancing "risk" on ten-year securities. Think about it: is our refinancing risk on two rollovers of a 10-year note so great that we should bear the additional cost of the 30-year? We think not.

Continued issuance of the 30-year is not consistent with our objective. Nor is it compelled by our constraints. The 30-year bond is not a necessary feature of efficient capital markets. With the cut backs in long-bond issuance carried out by my predecessors, it's role and liquidity had already been significantly impaired. Benchmark status had already shifted, several years ago, to the 10-year note.

Given our objective of maximizing our ability to finance at the lowest cost over time, and our desire to minimize the burden of our constraints on that objective, the decision to suspend the long-bond was relatively straight forward.

In the fourth quarter of last year, fixed-income markets experienced extraordinarily high levels of volatility, which only recently have abated somewhat. My hunch is that, several years from now that spike in volatility will be better understood as a reflection of an extraordinary conjunction of economic and financial events.

Essentially, financial markets were absorbing the slowdown in manufacturing and the rapid inventory adjustment, the bursting of the technology bubble, the dramatic swing in the federal government's short-run financing needs, and the immediate shock of 9-11 to confidence and economic activity and they also were beginning to anticipate an end to the economic slowdown. Those expectations of recovery help explain why the yield curve remained so steep - particularly at the short end. In addition, that steep curve provided an opportunity for millions of Americans to improve their personal cash flows. They did this by exercising the puts embedded in their fixed-term mortgages and refinancing at lower interest rates.

All of this took place in an environment of heightened risk aversion on the part of both our major financial institutions and speculative capital. This risk aversion diminished the pool of capital willing to step in to price the risk of both the likely and the unlikely outcomes.

As a society we have made a number of choices that reflect our collective desire to limit the variance in real economic outcomes - especially jobs and income. Particularly through federal fiscal policy, as I mentioned at the outset, but also through the provision of the put option embedded in conventional fixed-term mortgages and by the very structure and role of financial intermediation in America, we have made choices that result in the transmission of economic shocks to and through our financial system.

Having made these choices, we as a society are highly dependent upon the strength of our financial system to attenuate the transmission of shocks. But to do that, our financial system needs investors willing to commit capital in return for bearing risk.

But given the choices we've made perhaps, as a society, we should be somewhat less surprised to find so much volatility "going on" in our financial system.

Then again, we in the financial community need to do a better job explaining the importance of the transmission and attenuation of volatility throughout financial markets. Indeed, given its central importance in our financial system, we need to do a better job promoting the transparency, depth and resilience of the volatility market.

For my part, I thought it incumbent on me to do a better job of explaining how we see our role and what motivates our decisions in order to help you do your job of pricing the volatility that our behavior engenders. We cannot eliminate the volatility that debt management creates. We can, however, explain our objective clearly and identify the constraints under which we operate. Having done so, I hope we have at least taken a step towards greater understanding. Thank you.