I had a brainstorm in the Whole Foods store, and thought I would write it down. Currently, we should think of the Fed as operating in a regime much like the Bank of Canada's. This is a channel system, but one where the banking system holds a positive quantity of reserves overnight. This necessitates that the riskless rate on overnight interbank lending is equal to the interest rate on reserves held overnight. In its policy statement, the Fed mentions only its fed funds rate target, but the relevant policy rate is actually the interest rate on reserves, currently at 0.25%. As I discuss in my earlier entry, lending on the fed funds market has risk associated with it (which of course varies across borrowers), so it's hard to know what to make of the reported fed funds rate.
I mentioned in my earlier post the possibility of paying interest on reserves in the future at the T-bill rate. Scrap that idea. Let's think about how monetary policy works in a world where the riskless interbank overnight rate is identical to the interest rate on reserves. First, the Fed can choose what goes on the asset side of its balance sheet at will. T-bills, long-maturity Treasuries, mortgage-backed securities, agency securities - whatever. However, just like the Bank of Montreal circa 1925, it cannot determine the composition of its liabilities. Collectively, the nonfinancial and financial sectors determine how the quantity of outside money is split between currency and reserves held at Federal Reserve banks, much as note issuing banks were subject to the whims of their liability-holders.
Now, in this regime, what determines how much reserves are held relative to currency? The interest rate on reserves of course. In some sense, what is going on in this regime is that the Fed is doing both fiscal and monetary policy. Fiscal policy is determining the size of the balance sheet, and monetary policy is about setting the interest rate on reserves. Some talk I hear suggests that there are three instruments - i.e. add the fed funds rate to the mix. This is wrong, as the fed funds rate cannot be set independently of the other two instruments.
Next, how is monetary policy (the setting the interest rate on reserves) going to matter? It matters in exactly the same way that a conventional open market operation matters. A higher interest rate implies that the banks hold more reserves (essentially T-bills) and there is less currency in circulation. Perhaps the most amazing feature of the current environment is the large quantity of reserves that the banks are willing to hold at an interest rate of 0.25%. Obviously the Fed won't be able to sustain this and keep prices from rising without increasing the interest rate on reserves and/or reducing the size of its balance sheet.
Now, what about the Fed's proposed term deposit facility. The more I think about this idea, the more lame it seems. As I discussed in my earlier post, a bank's term deposit with the Fed will not satisfy reserve requirements and, more important, cannot be used for clearing and settlement. To induce banks to hold these term deposits, the Fed will have to offer an interest rate higher than the interest rate on reserves. Further, this will only make reserves scarce in the payments system, and increase the fed funds rate, for a given interest rate on reserves. Thus, the return on the Fed's portfolio falls, and efficiency is reduced. The Fed can make monetary policy as tight as it wants using the interest rate on reserves - the term deposit facility would only increase inefficiency.
Now, once we are back to "normal" and the size of the Fed's balance sheet has been reduced, we can go back to a channel system where the interest rate target is the fed funds rate, and there are no excess reserves in the system overnight. Of course, as any good New Monetarist knows, in this type of regime the fed funds rate is determined by the quantity (and maybe the composition) of assets on the Fed's balance sheet.