(Archived Content)
FROM THE OFFICE OF PUBLIC AFFAIRS
po-906
REMARKS BEFORE THE BOND MARKET ASSOCIATION LEGAL AND COMPLIANCE CONFERENCE
It is a pleasure to be here with you and to be back in New York. Thank you for inviting me.
There are just two things I want to talk to you about today: number one: the primary market for U.S. government securities and, number two, the secondary market for U.S. government securities.
I want to describe our efforts to improve the efficiency of the primary market by reducing the time it takes us to release auction results. The Treasury, the Bureau of Public Debt and our colleagues at the Federal Reserve Bank of New York are now on a mission to complete auction processing and release results consistently within two minutes. Achieving this will take some time and some changes for all of us. But our objective is clear and you can be judge of how we are doing.
I also want to touch on the importance of the efforts of your firms and your clients to ensure the continued efficiency of the secondary market by your vigilant oversight of the competitive dynamics of the financing market. The depth, liquidity and resilience of the secondary market for U.S. government securities are critical to our debt management strategy. But they are much more a function of the things that you do than they are of the things that we do.
Treasury auctions must be consistently brief: two minutes
The overarching objective for the management of the Treasury's marketable debt is to achieve the lowest borrowing cost, over time, for the federal government's financing needs. Although I have been at the Treasury for less than a year, I have been intimately involved in financing the government's marketable debt for almost a decade. While there are many things we have done to try to achieve the lowest borrowing cost for the taxpayer, for too long we have overlooked one of the simplest things we can do: namely, to reduce the period of time it takes us to announce auction results.
Processing bids and disseminating results more quickly will be a win-win situation for both investors and the Treasury. Shorter release times will reduce the period of time bidders are exposed to uncertainty as to whether and at what price they purchased Treasury securities. Reducing uncertainty will reduce risk for both investors and dealers. By reducing this risk, the Treasury will no longer need to compensate bidders for the implicit option premium associated with the extended period of uncertainty. This will lower the government's borrowing costs.
Considerable progress has been made in recent years. The Bureau of Public Debt, the New York Fed and the dealers and other submitters have worked together to make these improvements possible. In 1995 the average release time was 45 minutes. By 2000 average release times had been reduced to 27 minutes. And, as many of you may know, over recent months we released several auction results in less than 5 minutes by streamlining our treatment of questionable bids and by reducing the time the press has to turn the auction results into headlines. But we can do better.
To achieve the lowest borrowing costs, we must make the period of time between the auction close and the public release of results consistently brief. We will not squeeze out the implicit option premium we are now paying if we process most auctions in less than five minutes but still occasionally take 20 or 30 minutes to get results out. So our objective is a two-minute release and the variance we will tolerate is 30 seconds on either side of that objective. In the immediate future, our objective is to release auction results within six minutes, plus or minus 60 seconds.
We can meet the two-minute objective using existing technology, but to meet the two-minute mark consistently, we are all going to have to change our behavior to ensure that the process is driven by the rules rather than the exceptions to the rules. There will be changes in the submission of bids, the processing of bids and in how we get the results out.
Changes: your bid is your bid
For the bidding process, the most significant change - and the one that will require the most attention by dealers and investors - is also the simplest to understand: your bid is your bid.
For many years, we have been very forgiving of the mistakes and errors of submitters and bidders. We have been reviewing bids - eyeballing them for reasonableness - and giving submitters the chance to correct glaring errors that we found. The historical origin of our bid review lies before the dawn of computer systems when all auctions were manually processed. But to continue this review process in an automated world has come at the expense of getting our results out more quickly.
Beginning with our next refunding auctions, we will take bids as submitted and reject bids that fail to comply with our auction procedures. The onus will be entirely on submitters to ensure that their bids are accurate for themselves and their customers. This includes names, bidder ID's, par amounts, yields and net long positions.
Lest there be any confusion, let me state plainly: starting with the February refunding auctions, staff at Public Debt and the New York Fed will no longer proactively contact submitters to question bids or information submitted on tenders. Bids will be taken at face value and those that do not comply with our procedures and system edits will not be included in the auction.
From a systems' perspective, there will no longer be bids with incorrect par amounts or yields. All bids that satisfy the edits in the system will be accepted. Add an extra zero or mess up the big figure, and if it hits - it's yours. We will not delay the auction processing or release times to ensure that bids are correct and we will not accept any changes to bids that the system already accepted.
In addition, starting with the February refunding auctions, customer bids with an erroneous bidder ID number will not be included in the auction. It will be the responsibility of the bid submitter to work with the customer to ensure that this information is submitted properly.
You, in the compliance community, will need to continue to do your excellent work educating traders and customers about the auction process. But we are now going to make your jobs a little easier because there are going to be real consequences for bidding errors which should provide greater incentives to submit accurate bids.
That said, the consequences of submitting an erroneous bid that is accepted are manageable. Uniform price auctions ensure that all submitters will receive awards at a single, market price. The depth of the secondary market provides an adequate means of redress for those who may have erred in the par amount of their bid or had their bid rejected because of an unauthorized bidder ID or other problem.
Changes: reducing reliance on back-up telephone bids
Because bids are submitted through the communications links between your systems and ours, we also need to clarify how we will respond to emergency requests to submit bids via the telephone because of system malfunctions or failures.
In the past, extended auction release times have been a consequence of a submitter discovering, at the last minute, that their systems are not communicating with ours. Efforts have then been made to fix system problems and, occasionally, when fixes are impossible, to allow bids to be submitted over the telephone where adequate voice recognition exists. To shorten release times and, particularly, to reduce the variance, we must eliminate the possibility that these problems delay our auction results.
Eventually, our systems will be reliable enough that we will be able to eliminate backup telephone bids entirely. The next version of Treasury's automated auction system, TAAPSLink 2.0, scheduled for release later this year, will help us move in this direction
In the interim, beginning in February, should submitters experience legitimate systems problems, you will have to notify us at least ten minutes before an auction close and you will have to begin submitting bids at least five minutes before the auction close. We will enter all the telephone bids we can but only up to the close.
These are the immediate steps that we are taking to reduce release times. Three other areas where we will be working will take a little more time.
Further steps we will take to speed up our release time
First, we will continue to improve our technology, both hardware and software. We want to make submitting tenders more user friendly and we want to improve the speed and reliability of our processing.
Second, we are going to go back to the drawing board to rethink the application of the Net Long Position reporting requirement to see if we can find a way for compliance with this rule not to interfere with faster auction processing.
We appreciate your efforts to ensure compliance with our Net Long Position regulation. But the NLP rule, as currently applied, generates many tenders that are thrown into the questionable category by our system, requiring time-consuming manual review. Most NLP reporting errors tend to be procedural rather than substantive and almost all submitting firms would not be in violation of the 35 percent rule even if they received 100 percent of what they bid for. We do not now have a specific proposal, but we will be developing alternative approaches to enforcing the 35 percent rule that will not cost the taxpayers money by slowing down auction release times. I hope that we will be putting something out for public comment in the next few months.
The final area where we need to make changes to speed up our auction release times is in the public dissemination of our results. A surprising amount of the total time between an auction close and the release of results is currently taken up just with the process of getting the results out. We are going to reengineer this process completely - working with Public Debt's website, our communications links with submitters and the financial news services - to see what we can do to release results in as few seconds as possible.
These are the things we are going to be doing to improve the efficiency of the primary market. Some of these measures may seem strict in what I am asking of you and in what I am asking of Treasury, but they are vital to our achieving the objective of a two-minute auction.
In any one auction, accepting late bids, correcting bidder errors, or permitting backup telephone bids might lower our cost of borrowing on that day. But this comes at a long-term cost - of extreme variance in the duration of our release times - that we will no longer tolerate. We must look out for the taxpayer's long run interests.
The secondary market: you make it work
The efficiency of the secondary market for Treasury securities is something that we have less ability to influence by our direct actions. You and your firms - dealers and investors, risk managers and compliance officers - are the ones that can directly affect the depth and liquidity of the secondary market. We can write rules, we can implore and exhort you but, ultimately, your firms and your clients make this extraordinary marketplace function so well.
There is a tendency for us all to slip into the simplified habit of presuming that the liquidity of our secondary market is principally a function of how much debt we issue. There is, of course, an important and enduring truth to the idea that supply matters. But if that were all that mattered, there would be a great secondary market for grains of sand.
The liquidity and - as I like to focus on - the resilience of our secondary market is principally a function of the mechanisms that calibrate supply and demand. Most importantly, the smooth functioning of the financing market in general, and the market for specific issues in particular, play a vital role in the functioning of our secondary market.
As you will recall, last October we took the unprecedented step of holding an off-cycle reopening of the 10-year note. We took this action only after we concluded that normal market mechanisms were on the verge of failing.
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, you should not expect it and you certainly should not hope that we need to do it again.
We want to rely on you to reconcile the forces of supply and demand.
And for this reason, I ask you to redouble your efforts to self-regulate the efficient functioning of the repo market to ensure that there is a healthy competitive dynamic between the risks of holding long and the risks of holding short positions, between the risks of withholding collateral and the risks of failing to deliver.
Six years ago, when I was at the New York Fed, I sat down with Jerry Hawke, who was then the Under Secretary for Domestic Finance, to discuss the market surveillance of the repo market conducted by the New York Fed on behalf of the Treasury and the other regulators. Those discussions resulted in two speeches that I gave in October 1996 and in January 1997 in order to clarify the expectations the official sector had for behavior in the repo market following the 1991 Salomon Brothers' episode. (See http://www.newyorkfed.org/newsevents/index.html)
Today, Jerry is the Comptroller of the Currency and I am the Under Secretary and I feel just as strongly now as I did then about the importance of dealer firms self-policing the potential for extreme trading practices and squeezes of inpidual issues.
I am not going to repeat now everything that I said in those speeches - you have already been patient enough listening to me today.
I will suggest that anyone responsible for a government securities trading operation, whether as a manager, or in compliance, or in legal work, or in risk management, should take the time to read those two speeches. You should also read the remarks just given this past December by Dino Kos, my former colleague and successor at the New York Fed, at the Bond Market Association's recent Repo and Securities Lending Conference, which covers much of the same ground. (See http://www.newyorkfed.org/newsevents/index.html)
The reason I suggest the importance of reviewing this material is that I want this market to be self-policing to the greatest extent possible. I am still an optimist because I still believe that our interest in preserving the efficiency of the financing market is entirely consistent with good business practice on your part.
In conclusion, let me thank you for listening to me this afternoon and thank you for all that you do to make the U.S. government securities market the most efficient financial market in the world.