I was curious to take a look at the details behind the following story:
The Federal Reserve Board on Tuesday announced preliminary unaudited results indicating that the Reserve Banks provided for payments of approximately $46.1 billion of their estimated 2009 net income of $52.1 billion to the U.S. Treasury. This represents a $14.4 billion increase over the 2008 results.
The reported $46.1 billion figure is a 37% increase over the fiscal year 2008 value (as reported by the U.S. Treasury) and nearly double the average annual amount returned from the Fed to the Treasury over the last two decades. The Fed earned some income in 2009 from the limited liability corporations it created to manage some of the unconventional assets it acquired, and also reported some income from its new loans. Coincidentally, the combined income from these items turned out to be exactly equal to the various expenses that the Fed reported, with the result that the $46.1 billion returned to the Treasury is exactly equal to the $46.1 billion that the Fed reported it earned from the interest on its holdings of U.S. Treasury securities, agency debt, and mortgage-backed securities.
From Table 4 of the Monthly Treasury Statements we can also see when during the year the Fed turned these receipts over to the Treasury. The sum of the monthly totals that the Treasury reported it received from the Fed during 2009 comes to $43.8 billion, a little less than the $46.1 billion that the Fed claims it delivered to the Treasury. My guess is that the discrepancy results from the fact that the Treasury has netted out what the Fed describes as “amounts reimbursed by the U.S. Treasury and other entities for services the Reserve Banks provided as fiscal agents” whereas the Fed’s $46.1 billion is a gross total that excludes these offsetting payments from the Treasury back to the Fed. In terms of the monthly breakdown of the $43.8 billion number, 77% of it came during the second half of the year.
At the beginning of the year, when the Fed’s assets were concentrated in the various special new facilities, the Fed was returning less to the Treasury than it had on average over the previous 20 years. The big Treasury receipts began as the special lending facilities were wound down and replaced by outright purchases of Treasuries, agency debt, and MBS.
I have suggested that the right way to think about these operations is that the Fed acquired the funds for these operations by borrowing them from banks in the form of the huge expansion of interest-bearing reserves. In essence, the Fed is borrowing short through banks’ excess reserves and lending long through the MBS purchases, and profiting from the interest rate spread.
The ultimate rationale for such actions might be that private banks should be conducting this arbitrage on their own, but are prevented from doing so by ongoing financial difficulties. The spread between the MBS yield and the overnight rate can be broken down into two components, the first being a risk premium and the second being a term premium. We can get an idea about the size of the risk premium by looking at the difference between the yield on 30-year fixed mortgages and 10-year Treasury securities. This spread averaged 170 basis points over 1971-2009. It was well above that value in the last half of 2008 and first half of 2009, but currently stands at 134 basis points, somewhat below the historical average. But surely an objective observer would agree that at the moment mortgages are a riskier investment than they typically were over the previous 40 years. When you look at the current risk premium, it seems hard to motivate continuing Fed MBS purchases on the grounds that private banks are asking too big a compensation for mortgage-lending risk.
The second component of the MBS-overnight spread is the term premium, which we can measure by the difference between the yield on 10-year and 3-month Treasuries. This spread has averaged 176 basis points since 1982, but currently stands at 354 bp, about twice its historical average value.
Some might think the point of the MBS purchases was not so much to arbitrage an overly high long-term yield as to do something to stimulate the economy, since Fed purchases of short-term T-bills clearly accomplish nothing in the current near-zero interest rate setting. However, the difference between what the Fed has been doing in 2009 and true quantitative easing is that under the latter strategy, the Fed would be funding the MBS purchases with zero-interest money, with the intent of the operations indeed being to get that cash into circulation rather than trying to construct a device to persuade banks to hoard it.
The problem with the Fed’s preferred tactic– borrowing short and lending long– is the same as that facing anyone else who plays that game: are you sure you’re going to be able to continue to roll over the short term debt (i.e., persuade banks to keep holding a trillion in excess reserves) or liquidate your long position (i.e., sell a trillion in MBS back to the market without loss) when the playing field changes?
But at the moment, borrowing short and lending long proved to be a very profitable trade for 2009.