Wednesday, September 21, 2011

Operation Twist and Treasury Debt Management

As expected, the FOMC announced today that it will lengthen the maturity of its portfolio holdings of Treasury securities in an attempt to lower long-term interest rates.  No one is sure whether this will work.

One question that naturally arises is whether the Treasury Department will alter its debt management policy.  Under the current Administration, the Treasury has been lengthening the average maturity of the debt held by the public (which includes Federal Reserve Bank holdings).  Will Treasury continue that policy?  Probably the answer is yes, but, at the same time, Treasury will probably not make things more difficult for the Fed by taking advantage of its new initiative by selling more long-term debt than it would have otherwise.  The Financial Times published an article about this last week, misleadingly headlined “Treasury to accommodate Fed on ‘Twist.’”  In fact, the article only states that “the Treasury would be unlikely to respond to falling long-term interest rates with a sudden shift in the pattern of debt issuance, even though one of the Treasury’s strategic goals is to increase the average term of the US national debt.”  The article does not state that the Treasury would try to shorten the average maturity of the debt.

As for the original Operation Twist in the 1960s, one little remarked fact is that during this program, which was a joint Treasury/Fed initiative, the average maturity of the public debt increased from 1960 to 1963 (fiscal years). (See the table from the 1968 Report of the Secretary of the Treasury on page 74.)   In a short history of U.S. monetary policy posted on the New York Fed’s website, Treasury’s contribution is summarized as follows: “For a few months, the Treasury engaged in maturity exchanges with trust accounts and concentrated its cash offerings in shorter maturities.”  As for the Federal Reserve, it “participated with some reluctance and skepticism, but it did not see any great danger in experimenting with the new procedure.  It attempted to flatten the yield curve by purchasing Treasury notes and bonds while selling short-term Treasury securities.” (See page 39 of this document.)
While there have been some attempts to quantify the effect of the original Operation Twist, not much can be inferred from this experience, since Treasury’s contribution was minimal and eventually at cross purposes with the program.  Also, as Modigliani and Sutch in their much referenced article on Operation Twist point out, to the extent that there was a change in the term structure, the reason for this may have been attributable to increases in the interest rates that could be paid on time deposits at banks under the Federal Reserve’s Regulation Q which led to the creation of certificates of deposit (Franco Modigliani and Richard Sutch, “Innovations in Interest Rate Policy,” The American Economic Review, Vol. 56, March 1966, pp. 178-197.)  The effect on the yield curve of altering the supply of Treasury securities at various maturities is an unanswered question.  Assuming the Treasury does not undermine the Fed by its debt management decisions, the Fed is embarking on an experiment which will provide more data for those wishing to study this question.

Friday, September 16, 2011

A Note on Treasury Debt Management – Opportunistic vs. Regular and Predictable

At the August quarterly refunding, the Treasury asked its advisory committee on debt management, the Treasury Borrowing Advisory Committee (“TBAC”) to discuss the cost and benefits of extending the average maturity of the public debt and to provide a way to quantify those costs and benefits.  In its report to the Secretary, TBAC’s discussion of the maturity question was not particularly helpful:

“The presenters considered the total interest expense over time, the volatility of interest expense through time, as well as roll-over and liquidity risks.  The presentation highlights that longer dated term premiums appear elevated relative to the past.  That said, today there are uncertainties surrounding the long-term fiscal outlook, inflation expectations, and future borrowing needs.  A healthy discussion ensued amongst members as to whether or not the current long end premium was warranted.  While no definitive answer was reached, members felt that the current term structure of yields should not deter normal long-end issuance.  However, the Committee agreed that further analysis would be undertaken.”

More instructive, though, was the attached presentation (which can be found after the Treasury charts here).  The analysis seems to lean in favor of floating rate notes, which do not have rollover risk but are based on short-term rates.  The presentation also suggests that Treasury should issue more long-term debt when it thinks that there is a good chance that interest rates will rise more than the “term premium” embedded in long-term rates.  The presentation concludes that the term premium is relatively high but absolute rates are low.  The first bullet on the concluding slide states:  “The benefits of extension do not come for free.  Historical analysis suggests that shorter term funding has at many times been both cheaper and the volatility costs have not been high.”  They balance this statement with the following: 

“It is possible, however, that ‘this time is different’ because

o Nominal rates are much closer to the zero bound than previous periods

o Deficits are very  high historically and rising interest expense less acceptable

o Concentrated foreign ownership creates less reliable demand

o The benefits of funding attributable to being the reserve currency may be fading”

TBAC seems not to want to take a firm view on the maturity question.  This may be because the committee members have different views or because they do not want to say something with which senior Treasury officials may disagree.  What is interesting, though, is that the analysis focuses on current market conditions as the basis for deciding what maturities to issue.  This is a change from a long-term TBAC and Treasury view that debt management should be regular and predictable, and that Treasury should tap all maturity sectors regardless of current interest rates. The thinking was that the benefits to the Treasury of regularity and predictability outweighed any potential gain from trying to outsmart the market.  Treasury, it was thought, was just too big a borrower to do that successfully and consistently.  As one Assistant Secretary for Domestic Finance in the George H.W. Bush Administration put it: “It’s tough for an elephant to dance” (or something very close to that).

But the leadership of Treasury and the membership of TBAC have changed.  Treasury had been shortening the average maturity in the previous Administration; it is now lengthening it.  Part of the reason for this inconsistency is no doubt due to a weakening of the influence of the career staff on these types of issues.  During the late 1970s and until the mid 1980s, the key person making debt management decisions was a long-term career Treasury official, Francis X. Cavanaugh, who was adamant on regularity and predictability and a staunch believer in issuing 30-year bonds on a quarterly basis.  After he left Treasury, those views gradually became less accepted by a succession of political appointees.  Also, political appointees, who have a shorter-time horizon than most career staff, want to leave their mark on Treasury in the relatively short time they have to do it.  (I have, perhaps somewhat unfairly, in this regard remarked that political appointees just want to have fun.)  For those in charge of debt management, making changes is something they can do within the constraints of what the market will accept.

I am in the regular and predictable camp, but I do give the current Treasury credit for not surprising the market.  There is no clearly correct answer to the maturity question, and I am inclined to think that, within some reasonable and fairly broad parameters, it is not that important as long as the Treasury does not do something clearly stupid or rapid.  I do think the Treasury’s reputation took a hit in both the decision and the botched announcement of discontinuing the sale of 30-year bonds in the George W. Bush Administration.  However, while some people faced legal trouble because of their actions when they was a selective leak of the embargoed announcement [see clarification below], the decision was relatively quietly reversed after Peter Fisher left the Treasury, and  the memories of that episode have faded. 

Another interesting aspect of the maturity issue is the potential for Treasury and the Federal Reserve to be at odds.  The Federal Reserve apparently is considering a strategy of getting long-term rates down by reducing the supply of longer-term Treasuries in the market.  Rather than another quantitative easing program, the Fed is thinking of doing this by selling short-term Treasuries in its portfolio and using the proceeds to purchase long-term Treasuries.  Obviously, the Treasury could accomplish the same thing by selling more short-term Treasury securities and less long-term securities.  The Treasury, though, does not view debt management as a way to try to manipulate the term structure of interest rates but is focused on keeping its interest costs as low as possible “over time.” 

(There is an interesting recent blog post by someone using the name “Bond Girl” about this:  “Leave Operation Twist in the past.”  The author is not favor of the Fed trying to change the shape of the yield curve.)

Clarification:  With regard to the announcement of the cessation of 30-year bond issuance, the Treasury did not leak the news.  A consultant present at the press conference told a senior economist at a major investment banking firm about the announcement while the news was still embargoed.  The Treasury also broke its own embargo by posting the announcement on its website prior to the lifting of the embargo but after the consultant had told his client the news.  This was not one of Treasury's better days. 

Thursday, September 15, 2011

Solyndra, Interest Rates, and the Treasury’s Federal Financing Bank

The bankruptcy of Solyndra has drawn attention to a little known financing arm of the U.S. Treasury Department – the Federal Financing Bank or "FFB" – and the interest it charges to borrowers. The Treasury's Inspector General is reportedly investigating the role of the FFB in making this loan. I do not have any personal knowledge about the particular facts of the Solyndra loan, but, as way of background, here is some information about the FFB and how it sets its rates.

The FFB was established by the Federal Financing Bank Act of 1973. (I have been told that Paul Volcker, who was the Treasury's Under Secretary for Monetary Affairs at the time, is particularly proud of helping to create the FFB.) The purpose of the bank is to make more efficient the financing of loans which are 100 percent guaranteed by the federal government. Treasury securities carry cheaper interest rates than other paper sold in the market with 100 percent federal government backing, because they have greater liquidity and do not need to be explained to potential purchasers. Accordingly, the FFB advances the funds for 100% guaranteed loans and obtains the financing for this by borrowing from the Treasury. The Treasury in turn issues more securities than it otherwise would in order to fund FFB loans, though there is no attempt by Treasury to match any particular security with the FFB's borrowings.

The FFB used to charge a 12.5 basis point spread between its cost of borrowing from Treasury and the interest rate set on the loans it makes. According to the FFB's latest financial statements, any spread that is charged now for most loans goes to the agency which guaranteed the loans, not to the FFB, except if the FFB is not able to fund its administrative expenses associated with the loans. In that case, it "may require reimbursement from loan guarantors." The FFB does impose charges for the ability of borrowers to prepay loans and for forward interest rate commitments.

The methodology for setting interest rates on loans, both those at which the Treasury lends to the FFB and those at which the FFB lends to borrowers, is complicated. It does not involve simply reading the rate for a particular maturity off the Treasury yield curve, because Treasury securities provide a different stream of payments than most loans typically do. A Treasury note or bond pays interest semiannually and the entire principal amount at maturity. A typical loan is usually a series of level payments at specific intervals, each of which can be apportioned between interest and partial principal repayment.

The technical aspects of how these loans are priced are too involved to describe here precisely. I will endeavor here to provide a general explanation, which is unavoidably somewhat technical.

The FFB utilizes a model which effectively transforms the Treasury yield curve into a "theoretical spot rate" or zero-coupon curve. In other words, the assumption behind the Treasury yield curve is that the Treasury could sell a security at a given maturity at par with a coupon rate read off the curve for that maturity. A theoretical spot rate curve is a mathematical transformation of the Treasury yield curve to give the rates on zero-coupon notes or bonds that are consistent with the Treasury yield curve. For example, if Treasury can sell a note or bond at par with a 3% coupon, one could discount each coupon payment and the principal payment at maturity by the appropriate zero-coupon rates read off the theoretical spot rate curve in order to determine their present values. The sum of the present values of all the payments would equal the par value of the note or bond.

If a FFB borrower wants to make level payments for a fixed period at periodic intervals in order to obtain a loan for a specific amount, the FFB's model will determine the necessary size of those payments so that when discounted by theoretical zero-coupon rates, the present value equals the amount of the loan. The model will also calculate a "single equivalent rate" for the loan, which, using standard formulas, are used to determine the portions of each payment which constitute principal repayment and interest.

As for Solyndra, the latest FFB press release (July 2011) indicates that it had borrowed about $2.36 million from the FFB at a rate of 1.025%. This is the lowest rate of all the loans made by the FFB in June 2011 guaranteed by Department of Energy. However, the Solyndra loan matures in August 2016, while the other loans, except for one made to Solar Partners I, are for much longer terms and carry higher interest rates, which one would expect, given that long-term rates are higher than short-term rates. The Solar I loan, which matures in June 2014 carries an interest rate of 1.126%. Part of the difference between the rates on this loan and the Solyndra loan are no doubt due to the different pricing dates and the different payment frequencies for the loans – Solyndra is quarterly and Solar Partners I is semiannually. There may have been other details in the terms and conditions of the loans, but they cannot be ascertained from the press release.

It should be emphasized that for most of the loans it makes, the FFB is relying on the government guarantee. If the borrower defaults, it is the other agency which is on the hook to make the FFB whole, which has budget implications for that agency. The FFB, though, does face credit risk on the loans it has made to the U.S. Postal Service, and the financial troubles of that agency have been receiving some attention recently.

Also, investment bankers have not liked the FFB particularly, because its use means that they do not receive underwriting fees for non-Treasury securities that are 100% backed by the U.S. government. Of course, that is the point. Sometimes there are disagreements among various government agencies about the proper role of the FFB and its implications, and some 100% guaranteed loans are financed in the market.

In the Solyndra case, journalists and Congressional committees will no doubt continue to investigate. The role of the FFB may be interesting, but any investigation of why the loan was guaranteed and what analysis of the company's financial conditions was undertaken will need to focus elsewhere.

Update:  The press reports that the Department of Energy granted Solyndra a $535 million loan guarantee.   Bloomberg reports that the FFB made loans totaling $527 million to Solyndra.  I have no reason to doubt this, but it would be helpful if they had indicated from where this information came.  At its website, the FFB does not have information about its current portfolio itemized by borrower.  It does have information about its activity with specific borrowers for the month of June.  Consequently, it is not clear how much of the $535 million DOE guarantee was used to back FFB loans and what the interest rates were, except for the $2.36 million loan in June.  It may be somewhere on the web, but I have not been able to find it.  No doubt, further information will become available as investigations and bankruptcy court proceedings continue.  Since the FFB is something new for most of the press to cover, some news reports may be unclear, but perhaps the original source material will be available to those interested in this issue.

Monday, September 12, 2011

A Quick Note on the American Jobs Act and Entitlements

Liberals in general appeared to be pleasantly surprised by President Obama's speech to Congress last Thursday in which he proposed a $447 billion package of tax cuts and spending increases designed to spur job creation. For example, while Paul Krugman had some reservations about the President's proposal (mostly concerning the effectiveness of tax cuts), he wrote in his column that he "was favorably surprised by the new Obama jobs plan, which is significantly bolder and better than I expected." Liberals, however, may be less happy about how the political need "to pay" for this package may impact long-term deficit reduction proposals that will be produced by the new super committee created by the debt limit legislation.

According to media reports, the White House has outlined some tax proposals to cover the $447 billion package. However, effectively the $447 billion will be added to the amount the super committee has to find in order to prevent automatic spending cuts. This implies that entitlement programs, such as Medicare, Medicaid, and Social Security are likely to be even more inviting prospects for spending cuts, given the larger amount of deficit reduction measures that need to be found in order "to pay for" whatever jobs legislation eventually is enacted.

The American Jobs Act proposal seems to have a dual purpose, to jump start the economy and to achieve entitlement reform. In other words, it is an attempt by the Administration to reach an elusive "grand bargain" that was not achievable during the debt limit negotiations.

Some conservatives, donning the mantle of being "responsible," are saying that this is all to the good and the Republicans should take a deal of short-term stimulus for "serious" entitlement reform. (For example, see this column by David Brooks.) When it becomes clearer that President Obama is trading temporary spending increases and payroll tax cuts for permanent reductions in key New Deal and Great Society programs, liberals are likely to feel disappointed once again. While some elements of the Republican Party have implausibly attempted to paint Obama as some kind of socialist, liberal Democrats are realizing that the President is not as liberal as they once thought.

Federal Reserve Options and Treasury Debt Management

Last Wednesday (September 7), the Washington Post reported that one of the options the Federal Reserve was considering to bolster economic growth involved selling shorter-term Treasury securities in its portfolio and purchasing longer-term Treasuries. This is in contrast to its QE2 program, in which the Fed created bank reserves to pay for long-term Treasuries.

Last October, I wrote about the potential conflict between the Fed and the Treasury on attempts to manipulate the shape of the yield curve ("Monetary Policy and Treasury Debt Management -- A New Operation Twist?"). What I said then is relevant now. Moreover, the policy reportedly under consideration at the Fed could potentially change the maturity structure of the public debt held outside of the federal government trust funds and the Fed even more than QE2, because the new proposal entails the Fed adding to the supply of short-term debt.

During the George W. Bush Administration, the Treasury shortened the average maturity of the public debt, which culminated in a badly introduced policy of stopping the sale of 30-year bonds, which was reversed after the Treasury official responsible for this decision, Peter Fisher, left the Bush Administration. During the Obama Administration, the Treasury has been lengthening the average maturity.

For reasons that are not entirely clear, the Fed has for a long time argued that the Treasury should sell less long-term debt. For many years, the Treasury ignored this advice, but began heeding it in the Clinton Administration by reducing the amount of 30-year bonds, which used to provide the benchmark long-term interest rate to the market but has been supplanted in this role by the 10-year note. Peter Fisher of the Bush Administration came from the New York Fed and was unusually solicitous of debt management advice from the staff of the Federal Reserve Board. While Secretary Geithner was President of the New York Fed, the Treasury has reversed course on the debt maturity issue. Since Fed officials do not usually speak publicly about this type of Treasury debt management issue, there is no direct evidence about what they think of Treasury debt management policy. However, the idea of altering the maturity structure of the public debt in the market would seem to indicate that they believe that Treasury has been issuing too much long-term debt.

The Treasury, if it wanted to, could overwhelm the Fed through its enormous financing operations. One should recall in this respect that the supply of new debt is not only that due to financing the deficit but also to refinancing debt that comes due. The Treasury would be unlikely to issue more long-term debt in order to undo the Fed's policy, should it be implemented. The responsible Treasury officials would probably view the Treasury's liability portfolio to include Fed holdings, even though the interest Treasury pays on that debt is returned to the Treasury after subtracting Fed operational expenses (including the 25 basis points of interest paid on both required and excess reserves).

This highlights a gray area between the Treasury's and the Fed's responsibilities. It is not clear how much discussion between Treasury and Fed officials there has been over this issue. I would guess that there has been some and that the Treasury has indicted that it would not object if the Federal Reserve Open Market Committee decides to alter the maturity structure of the debt in the market. Given the Treasury's responsibility for debt management and its ability to work at cross purposes to the Fed through its debt management operations, I would hope that the Fed does not feel it can do whatever it wants on an issue like this without consulting the Treasury.