Sunday, November 7, 2010

Could QE2 Cause the Fed to Go Broke?

The Fed's new program of quantitative easing, QE2, once again raises an old question: Can central banks go broke? Conventional analysis, aptly summarized by Willem Buiter in a 2008 report, says no, or at least, hardly ever. However, when we look closely, the conventional analysis is not altogether reassuring. Although the Fed most assuredly is not going to go broke, preventing that from happening could raise difficult political issues and perhaps even threaten the Fed's independence.

We can start by noting that the Fed, like most central central banks, is rather thinly capitalized. As of November 3, 2010, it had capital of some $56 billion, about 2.5% of its assets of $2,303 billion. By comparison, Bank of America, with approximately the same total assets, had 7.8% Tier 1 common equity at the end of 2009. If the Fed were a commercial bank, its financial condition would not be dire, but it would be on the watch list.

Of course, the Fed is not a commercial bank. As the conventional analysis is quick to point out, the unique nature of its assets and liabilities normally allows it to operate safely with just a sliver of capital. Normally, the Fed's assets have consisted largely of short-term Treasury securities, which are as close to risk-free as you can get. As for liabilities, as recently as the end of 2007, 90% of them consisted of Federal Reserve currency. Currency is a truly marvelous thing to have on the right-hand side of your balance sheet, since it is neither interest-bearing nor redeemable. With a assets and liabilities like that, who needs capital?

Since 2008, however, alterations in the Fed's balance sheet have undermined the conventional analysis to a certain extent. First, the nature of assets has changed, and continues to change. The Fed's all-Treasury asset portfolio is only a memory. It now holds more than a trillion dollars worth of mortgage backed securities that are neither very liquid nor risk-free. In addition, QE2 is in the process of lengthening the maturity of the Fed's Treasury portfolio, so that while there is still no credit risk, there is growing exposure to market risk in the event of a rise in interest rates.

On the liability side, nonredeemable monetary liabilities still predominate, but the composition of the monetary base has changed. More than half of the base (as opposed to less than 10% three years ago) now consists of reserve deposits of commercial banks. Reserves are no longer interest free. True, as of 2009, interest expense on reserve deposits was less than 4% of the Fed's net interest income, but that is up from zero. And keep in mind that while the rate paid on reserve deposits is now just 0.25%, a potential increase in that rate looms as part of the Fed's exit strategy from its current expansionary policy stance. Another potential exit strategy tool, reverse repurchase agreements, also comes with interest cost attached.

That, shaped by earlier rounds of quantitative easing, is the starting point from which QE2 is being launched. Suppose that at first QE2 has little impact on the economy, but then, about the time the Fed's balance sheet hits the $3 trillion mark, inflation expectations and interest rates begin to rise. Perhaps they rise sharply as everyone tries to bail out of Treasuries before prices collapse. As promised, the Fed counters by implementing its exit strategy, selling bonds at a loss, using reverse repos on a large scale, and raising interest rates on excess reserves. The Fed's net income would certainly decrease, and it is far from impossible that its capital could drop below zero. What then?

First, it should be made clear that even if the Fed slipped into balance-sheet insolvency (negative capital), that would not bring about equitable insolvency (inability to meet financial obligations as they fall due). Because of the nonredeemable character of its monetary liabilities, and because both its liabilities and assets are denominated in dollars, any kind of run on the Fed is absolutely impossible. Beyond interest on reserves and reverse repos, the Fed still would have to meet some six or seven billion dollars a year in operating expenses and obligatory dividends to member banks, but even if its net interest income were much reduced from the $50-odd billion it will earn in 2010, it could probably cover these.

Still, a position of negative capital would be uncomfortable even if the Fed were able to keep up with its current obligations. Recapitalization would clearly be desirable. But just how could it be accomplished?

Recapitalization would be complicated by the Fed's odd legal status as a joint-stock entity that is "owned" by private commercial banks, yet is in every functional sense a part of the federal government. The only conceivable entity that could recapitalize the Fed is the Treasury, but this would be no ordinary capital injection. For commercial banks, a capital injection means a swap of good assets for equity, but the Fed could not just issue new common or preferred shares to the Treasury, at least not without a revision of its charter. Instead, a recapitalization would have to take the form of an outright grant, in which the Fed transferred tens or hundreds of billions of dollars in newly issued bonds to the Fed completely gratis. It is hard to see how that could be done without an act of Congress--and would Congress in its current mood approve this mother of all bailouts?

Let me emphasize this: The Fed is NOT a private corporation in any ordinary sense of the word. If the Treasury were to gift the Fed with a $100 billion capital grant, that would NOT amount to putting it in the pockets of the Rothschilds, whatever you might read to the contrary on the internet. But, can you guarantee that all those paranoid myths about the Fed would not be raised in Congressional debate or on talk radio? I cannot make that guarantee, and for that reason I cannot guarantee quick passage of the Treasury Asset Recapitalization Package of 20**, or whatever they might call it. Whatever the name, it would be called TARP II and it would be controversial. It would be so controversial that in return for passage, Congress might insist on new audit or oversight authority, something already high on the agenda of certain members.

So, what is the bottom line? Could the Fed go broke if QE2 creates a bond bubble that suddenly bursts in a surge of inflationary expectations? In fact, it actually could become insolvent in the balance sheet sense. Presumably, it could not become insolvent in the equitable sense. But we cannot rule out the emergence of a situation from which the Fed could be extracted only at the cost of a high degree of political discomfort and perhaps a loss of independence.

Follow this link to view or download a short slide show with data from the Fed's balance sheets and further analysis.


  1. The Fed could also borrow against its future seigniorage revenues. The monetary base is (say) 10% of US GDP, growing at (say) 4% per year in nominal terms, discounted back at (say) 5%. So annual seigniorage is 0.4% of GDP, growing at 4%. The present value of that is 40% of GDP. It's big.

  2. Nick-- Many central banks, notably the People's Bank of China, issue their own bonds or bills in that way. Selling Fed bills would be a good tool for a future exit strategy, since it would allow the Fed to exchange monetary liabilities (reserves and currency) for nonmonetary liabilities (bills). However, there are two problems here.

    First, selling debt could provide cash for the Fed to meet current obligations, but it would not increase its capital.

    Second, the Fed apparently lacks legal authority to issue bills. It reported discussed the option with Congress in 2008, but was turned down. (See this Bloomberg piece, for example. ) Some people seem to think it could invoke an "incidental powers" clause in its charter to issue bills, but it might get roughly slapped down by the courts and Congress if it tried. Congress guards its powers to limit federal debt rather jealously (viz the annual fight over raising the Treasury's debt limit) and would not be eager to let the Fed evade the ceiling through the back door.

  3. Ed, I'm late to the game but this is a very good post you've written.

    One other facet of the question of central bank capitalization is its influence on central bank clearing house obligations. Most central banks run their nations' respective payments clearing operations. A good clearing house must be well-capitalized in order to gain the confidence of participants. A central bank moving towards insolvency loses credibility not only as the monopoly issuer of currency, but also as the monopoly manager of its payments clearing system. The macro effects of this could be panic, and/or some movement towards competing clearing systems, whether these be informal barter-like systems or more sophisticated private alternatives.