Yves Smith points us to a couple of pieces discussing the Fed's "balance sheet constraint", the notion that the central bank may run out of treasury securities to exchange, whether temporarily or permanently, for the questionable securities held by private banks. This asset swap has emerged as the Fed's core response to the current crisis, and is the essence of what James Hamilton referred to as monetary policy on the asset side of the balance sheet. In an excellent summary, Greg Ip describes the various options the Fed would have if it were to run low on Treasuries.
Fundamentally, the Fed would have two options: It could increase the size of its balance sheet by issuing cash, which would require sacrificing its target Federal Funds rate target and letting that rate drop to zero. This option is referred to in the trade as "quantitative easing", but that's just a fancy term for printing money and tolerating any inflation that results. Alternatively, the Fed could expand its balance sheet by borrowing from someone else — from the US Treasury, from banks with excess cash, or from the public directly. This would permit the Fed to increase the scale of its asset swaps without sacrificing its ability to conduct ordinary monetary policy.
If you want to understand the details, do read Ip's piece. The Fed's "balance sheet constraint" is not a hard limit. The Fed can circumvent it. But that doesn't mean that the size of the Fed's balance sheet is not important. Consider this, from Ip (emphasis mine):
The easiest would be to ask Treasury to issue more debt than it needs to fund government operations. As investors pay for the bonds, their cash moves from bank reserve accounts at the Fed to Treasury accounts at the Fed. The Treasury would allow the money to remain there, rather than disbursing it or shifting it to commercial banks who, unlike the Fed, pay interest. Because the shift of cash out of reserve accounts leads to a shortage of reserves, it puts upward pressure on the federal funds rate. To offset that, the Fed would enter the open market and purchase Treasurys (or some other asset), replenishing banks’ reserve accounts. The net result is that the Fed’s assets and liabilities have both grown but reserves and the federal funds rate are unaffected. This wouldn’t cost Treasury anything so long as it doesn’t bump up against the statutory debt limit. The loss of interest on its cash deposits at the Fed would be roughly offset by the additional income the Fed pays Treasury each year from the interest on its bond holdings.
It's only true that this operation doesn't cost the Treasury anything if what the Fed buys with the excess cash pays as much as the Treasury's cost of borrowing, and there is no loss of principal. But if the Fed uses the cash (directly or indirectly) to buy or lend against market-shunned securities, then the Treasury is only made whole if those securities perform, or the loans against them are repaid. If the market is irrationally shunning these securities, then the Treasury will eventually break even. But if the securities turn out to be worth less than what the Fed lends or pays, taxpayers might be forced to eat the loss.
Fundamentally, the Fed's balance sheet constraint is and should be a political constraint. The size of the Fed's balance sheet defines how much capital taxpayers and holders of currency are making available to the Fed to do whatever it is it's doing. Whether Fed's balance sheet should be expanded is an investment decision — should the public throw more money at the project that the Fed is undertaking? There's a real downside — losses by the Fed will eventually be borne either by taxpayers or by owners of dollar denominated assets (which means especially workers with little bargaining power, whose wages are negotiated in nominal dollars and would not rise with inflation). But bearing those risks may be less damaging than the harm that would result from turmoil in the financial system if the Fed loses its capacity to act.
I don't know whether expanding the Fed's balance sheet is a good idea, if it comes to that. There are risks and benefits associated with the Fed's proposed use of funds, and reasonable people can come to very different judgements. What I do know is that a decision to expand the Fed's balance sheet ought not be treated as technocratic monetary policy. However funds are raised, their repayment would be guaranteed, so all downside risk would be borne by the public. Expanding the Fed's balance sheet would represent a sizable investment of the public's wealth, and the public ought have as much say over that decision as over any other investment of public money.
Update: Now here's some creative thinking! These so called "reverse MBS swaps", under which the Fed would refill their stock of Treasuries by swapping back iffy securities wrapped with a Fed guarantee, would have no direct balance-sheet impact whatsoever, and if repeated would provide the Fed with a potentially infinite supply of Treasury securities to swap! Of course, the proposal is simply a scheme to create off-balance-sheet liabilities in order to evade what might be on-balance-sheet limits. Wow.
I frequently marvel about how, in order to respond to the credit crisis, the Fed as well FHLB, Fannie, and Freddie, are doing precisely what got private actors into their messes in the first place. Off-balance-sheet liabilities are a logical next step.
(This was reported in Greg Ip's piece, but somehow I didn't grok the implications until reading Lou Crandall's description , "[The reverse MBS swap] is sufficiently exotic that it might sidestep some of the traditional legal issues." That kind of line is a spur to really think things through!)
- 6-Apr-2008, 4:20 p.m. EDT: Added update re "reverse MBS swaps".