The Federal Financial Triangle

What would it mean for the world’s principal central bank to have negative net worth?

By Alex J. Pollock

As odd as it may seem to us now, under the National Banking Act from 1863 to 1913, local banks with national charters were the official issuers of U.S. currency, and the government had no central bank. Hundreds of national banks in towns and cities all across the country were issuing dollar bills. They were replaced in this essential role by the Federal Reserve Banks under the Federal Reserve Act of 1913.

Consistent with their responsibility as the issuers of U.S. paper money, national banks were prohibited during this period from making any real estate loans at all. Today, in contrast, the Federal Reserve is a huge investor in real estate loans. It owns over $1.1 trillion of them — and keeps buying more — in the form of mortgage-backed securities (MBS). This would have greatly surprised and highly displeased the authors of the Federal Reserve Act.

But don’t worry about the credit risk of these mortgage loans: the MBS the Fed keeps buying at the top of the market are guaranteed by Fannie Mae and Freddie Mac. Whoops: Fannie and Freddie both went completely broke, suffering staggering aggregate losses of $246 billion, which wiped out all their capital and a lot more.

But not to worry: Fannie and Freddie were given new capital — $187 billion of it — by the U.S. Treasury, in order to bring their capital up to zero. The government effectively now owns and runs both of them; they are no longer “government-sponsored enterprises,” but simply part of the government. Since the Treasury keeps their capital at zero, they operate with infinite leverage. But not to worry: if things go wrong again, they can always get more money from the Treasury.

But where does the Treasury get its money? Well, to a very significant extent, it gets it from the Federal Reserve, which has so far amassed more than $1.8 trillion of Treasury debt, and keeps on buying — again at the top of the market.

This all makes a most interesting triangle of government finance, as shown in Figure 1. Thinking about it, my brother, a Swiss private banker, observed, “Just about what John Law built in France in 1716-1720” — referring to the notorious paper money theorist and government banking practitioner, who inflated the infamous “Mississippi Bubble” of his day.

The Fed, by buying long-term MBS and long-term Treasury debt, especially by buying them at the top of the market — a top which its own purchasing pressure is intentionally creating — is concentrating massive interest rate risk on its own balance sheet. By “the Fed’s” balance sheet, we actually mean the aggregate balance sheet of the 12 Federal Reserve Banks. This consolidated balance sheet has $3.3 trillion in total assets and $55 billion in equity, for leverage of a heady 60 times and a capital ratio of a paltry 1.7 percent.

A rise in interest rates from their historic and Fed-manipulated lows is inevitable at some point. To trigger this, the Fed would not have to start selling, only to stop buying. How vulnerable is the balance sheet of the Fed to interest rate risk? We can estimate the market value impact on the Fed of a 2 percent increase in interest rates, which is a common benchmark. Of course, it is very easy to imagine — or forecast — rates rising by a lot more than that in the medium term.

When considering the Fed’s $1.1 trillion in MBS (to which it is adding $40 billion a month), a reasonable estimate is that a 2 percent rise in interest rates would reduce the market value of long-term fixed rate MBS by about 15 percent. On its $1.1 trillion portfolio, this would be a hit of about $169 billion to the value of the Fed’s assets.

Turning to the $1.8 trillion in longer-term Treasury securities, a recent study by Fed economists (“The Federal Reserve’s Balance Sheet and Earnings — A Primer and Projections,” January, 2013) shows the weighted average maturity of the Treasury portfolio at about ten years. The study suggests that the Fed might run major net losses from 2015 to 2020, when it would have to sell its by-then-underwater securities. The Fed has also bought $80 billion in federal agency debentures — added to the Treasuries, that makes $1.9 trillion.

The Fed’s portfolio of Treasuries and agencies has a mix of various maturities, of course. Based on the maturity breakdown in the Fed’s December 31, 2012, quarterly financial report, one might guess its duration overall is about 7.5 years. A 2 percent rise in rates would thus reduce its market value by about $285 billion.

So a reasonable estimate is that, adding the MBS and Treasuries together, a 2 percent rise in interest rates would reduce the market value of the Fed’s total long-term investments, which are unhedged, by about $454 billion. With greater increases in interest rates, the losses would be correspondingly greater, possibly much greater. As of December 2012, the Fed’s fair value disclosure seems to suggest an unrealized gain on its investments of $193 billion. With a 2 percent rise in interest rates, the consequent reduction in value of $454 billion, net of that $193 billion, would leave a loss of $261 billion. This is almost five times the Fed’s capital of $55 billion — or a marked-to-market net worth of a negative $206 billion. This would become realized if the underwater investments were sold.

If (or when) this happens, what would it mean for the world’s principal central bank to have negative net worth? That is not clear — some people argue it would not matter as long as the Fed can keep printing currency. But the Fed obviously does not want to find out what the market, and political, reaction to such an outcome might be.

Thus the Fed has fixed its accounting so that any net losses will not reduce its reported capital. If with capital of $55 billion, the Fed had realized net losses of $55 billion, you might think its capital had become zero, as it would for anyone else in the world. But not to worry: as far as Fed accounting goes, you would be wrong.

The Fed controls its own accounting standards and, in 2011, changed its accounting so that even with a net loss of $55 billion, its capital would still be reported as $55 billion. If it lost $100 billion, its capital would still be $55 billion. If it lost $200 or $300 billion, its capital would still be $55 billion. Get it? Fortunately for the Fed, at the apex of the federal financial triangle, its own accounting policy has no pesky interference from say, the Securities and Exchange Commission.

An important consideration is whether the federal financial triangle should be consolidated into a single set of government books, with all the obscuring intragovernmental transactions becoming consolidating eliminations. It is a reasonable argument that we should look at the triangle this way. What do we have then? Net: the government issuing paper money and buying real estate loans.