FROM THE OFFICE OF PUBLIC AFFAIRS

When I was invited to participate in today’s panel on interest rates, I was as surprised as anyone that the word “yes” came out of my mouth. It is not a surprising invitation, given how important interest rates are to me in meeting my job objective of providing Secretary O’Neill with advice on how to best finance the Government.  After all, our number one objective is to finance the Government at the lowest cost over time.  What you may not be aware of, is how little a role interest rates play in the advice I provide.

We issue debt regularly and in predictable quantities, rather than opportunistically. A consequence of regular and predictable issuance is that we are not in a position to tailor debt issuance to interest rates.  And as I just said, not only does this make sense for Treasury but we really have no other choice. 

If Treasury officials were to alter issuance to take advantage of interest rate fluctuations, they would not necessarily lower borrowing costs – any price concessions from exercising market power are likely to be more than offset by the your superior resources devoted to understanding interest rate movements and modeling our behavior. 

Aside from lacking resources -- the office of market finance has no more than four people devoted to debt policy -- it is difficult to think of a multi trillion-dollar-a-year annual issuer of debt as nimble.  The shear scale of our operations dictates a high degree of regularity in issuance.  What we have done at Treasury is turn some degree of necessity into a high degree of commitment.   So we don’t hold snap unscheduled auctions for a given maturity when yields appear low, and we don’t even take the yield curve into account when we allocate how much to raise by different maturities.  Our issuance calendar is well known to every trader or investor involved in our market.  The auction calendar is published months in advance.  I have no doubt that the trader in all of us believes that he or she could “beat” the market by opportunistically managing treasury’s debt issuance.  But once we start to try and capture interest rate moves on the margin we will have become traders and not debt managers.  The risks associated with our own judgments are insurmountable.  We can’t trade our way out of a bad position.

Underlying this regularity of issuance, however, is an entity managing the world’s largest cash flows.  Over the past year, Treasury received and spent roughly $2 trillion. 
As most of you know expenditures and receipts do not coincide – we make large entitlement payments at the beginning of months and receive cash sporadically throughout months, with receipts lumped unevenly around only a few tax dates.  Part of my job is to figure out how to efficiently manage uneven cash flows with regular debt issuance.  Part of the consequence of my advice is huge cash swings in Treasury accounts held in the banking system.

It is not unusual for our cash accounts to swing by more than $50 billion over the space of a few days.  By contrast, if we were to change issuance by more than a few billion from one auction to the next, we would surprise market participants.  Our situation is further constrained by Treasury’s strict collateral requirements with the banks that hold our cash balances.  Constraints, however, do not alter our objective:  the better we can manage our cash balances, the better we can serve the taxpayer. 

In managing these balances, we face a significant challenge in how we respond to the uncertainty associated with our cash flows.  We know that we will have large seasonal swings in our cash balances and we know when those swings will occur, but we don’t know how big they will be.  Our ability to respond to these fluctuations is limited by our auction calendar – our most flexible tool, weekly bill issuance, settles once a week and our longer term securities only settle on a monthly or quarterly basis. 

Uncertainty about the future also constrains debt management planning.  We have to be prepared to finance either sustained surpluses or deficits.  The range of potential fiscal outcomes is pretty remarkable – CBO measures this range over a five-year horizon in trillions of dollars.  My job is to ensure that, no matter which outcome materializes, Treasury will be able to finance the Government’s needs without forgoing its commitment to regular issuance.  Because we are regular, because we want to prepare the market, our auction calendar must be able to adapt to a variety of outcomes.

The average miss on the current year of forecasting, and this includes street economists, CBO and OMB – and I repeat – the average miss on the current year after four months of actuals is $75B.  That’s pretty amazing.  This makes the job of debt manager tricky because we already saw how limited our options can be.

As a result of this uncertainty, we cannot “time” expectations.  That is, debt management cannot be predicated on daily, weekly or monthly concerns about what may or may not be a change in the economy.  Changes in debt management are based on a deliberative process after consultation with market participants.  We have a formalized, transparent process around our quarterly borrowing so that we can solicit your views before we make issuance changes and you can read about the factors that influence our decisions.  The process deliberately prohibits decisions based on short-term expectations.  This slow and predictable behavior is frustrating to many of you, but once again, if our decisions are wrong, we can’t call up dealers and get out of positions.

While we do not, and cannot, base our decisions on current interest rates, we think continuously about how our issuance pattern influences our cost of borrowing.  This thinking is motivated by Secretary O’Neill who has challenged us to improve all facets of our operations.  Faster auction times, increased transparency, better systems, and improved analytics are helping us achieve this goal, but we seek further improvements.

On the operations side, I invite you to test our systems by participating directly in our auctions.  Our current distribution system works exceptionally well, but for those of you who have or considered participating in an auction, Treasury now offers safe (word) point and click capability.

On the debt policy side, I encourage you to think about our issuance and how we can reduce our interest costs.  In October, we solicited views from our private sector advisory committee on how to measure debt management performance.  Going forward, we will continue to seek views and formulate methodologies that will provide better quantification of what we do and how we do it.   
 
I will conclude with a couple of observations about our issuance from a historical perspective.  Debt markets in the U.S. have grown more rapidly over the past two decades than Treasuries.  I view this as good for two reasons. One, it is a sign that the government is playing a smaller role in financial markets. Two, from the perspective of a debt issuer, our smaller portion of supply means that we can issue more cheaply and that our responses to the inevitable forecasting errors we face are more easily absorbed by market participants.

In spite of this reduced position, I recognize that Treasuries act almost as a currency for a wide range of transactions and that Treasuries play a special role in asset allocations. I expect that these roles will grow over time as you, the users of Treasuries, continue to develop faster and more efficient financing mechanisms. 

I would like to thank you for your time and would be happy to take any questions.

Chart Package