Tuesday, February 10, 2009

Is U.S. Monetary Policy Really Tapped Out?

Many observers have been making the case for fiscal policy by arguing that conventional U.S. monetary policy has exhausted all of its ammunition. For example, Brad DeLong says:
...We need a big fiscal boost program because monetary policy is already tapped out--Treasury interest rates are at zero--and employment losses are about to be bigger than in any previous recession since the Great Depression.
Other observers go on to say that even unconventional U.S. monetary policy, such as the Fed's quantitative easing, has its limits. Representing this view is Paul Krugman:
Yes, there are other things the Fed could do — and it’s doing them, on an awesome scale. But they’re controversial, precisely because, unlike conventional monetary policy, they involve picking and choosing among potentially risky investments. And there’s a much stronger case for fiscal policy than in normal times, because we don’t know how well these unconventional measures will work.
So conventional monetary policy is not working and the efficacy of unconventional monetary policy is uncertain. Maybe so, but I have a few questions. First, how do we know that unconventional monetary policy is not working? On the credit crunch front there is evidence that the Fed's policies are working to some degree as pointed out at macroblog. Also, monetary policy's effect on the real economy typically occurs with a lag so it seems premature to pass judgment here. Second, are we even sure that unconventional monetary policy is being fully exploited? I say this because of the Fed's policy of paying interest on excess reserves. This policy seems highly counterproductive given the state of the economy. What would happen if the Fed dropped this policy and did not attempt to sterilize these reserves? I suspect we would not see pictures like this one (click on figure to enlarge):

For those who are concerned that such a move would let the inflation genie out of the bottle I would (1) point you to this troubling figure that suggests deflation is more of a threat and (2) ask you to read Nick Rowe's discussions here and here. It is possible that banks would continue to hang on to the excess reserves in the absence of this policy, but given the improvements in credit markets it seems that at least some of these excess reserves would be put to good use. So I ask again is U.S. monetary policy really tapped out?

Update: See JKH's discussion in the comments section. Among other things, he explains that contrary to my assertion above the Fed's policy of paying interest on excess reserves is not consequential to the current excess reserve buildup.


  1. “What would happen if the Fed dropped this policy and did not attempt to sterilize these reserves? I suspect we would not see pictures like this one (click on figure to enlarge)”

    Why should one suspect this?

    The Fed determines the level of excess reserves. Member banks have no direct effect on the level of system reserves, whatever their chosen velocity of reserve deployment.

    Are you assuming the Fed would drain these excess reserves, as a result of banks choosing to “deploy” excess reserves (i.e. increasing the velocity of reserves), in the event they earned no interest on them?

    I.e. are you assuming that the Fed would actually tighten its existing quantitative policy if it did not pay interest on reserves and the banks became more active in accumulating assets?

    (The Fed would actually have to sterilize existing excess reserves in order to cause a reduction in the system reserves level.)

    This seems a bit of a stretch in a deflationary environment. Banks lend as a function of capital more than reserves. Are you assuming that the banks would acquire that many more zero risk weighted assets such as treasury bills?

    The Fed’s primary reason for this excess reserve position is to fund its qualitative (credit) easing program on the other side of the balance sheet. Draining excess reserves would be self-defeating in this regard, as an alternative source of funding would be required. They’ve already announced the winding down of extraordinary treasury funding – which is the reason why they’ve expanded reserves as a source of funding in the first place. Paying interest on reserves is a structural change that allows them to set a floor on the funds rate as per their target.

    Moreover, banks are earning near zero interest on reserves as it is. Do you really think dropping the interest earned would have more than a negligible effect on reserve velocity in this environment?

    (I’m not sure what the rate is now, but it can’t be more than the target funds rate range ceiling of 25 basis points.)

  2. JKH:

    After thinking this through some more, I believe you are correct that the effect of removing the payment of interest on excess reserves would be negligible. The rate they are paying is so low it has to be inconsquential. Thanks for helping me think more clearly on this issue.

    So let me throw you a question slightly off topic. What danger is there in the Fed targeting the long end of the yield curve? I have heard that it would be harder to reverse such actions, but if it were limited to highly liquid long-term Treasuries this shouldn't be a concern. Any thougths

  3. David,

    I would break your question into several parts.

    First, the world has changed now that the Fed has started to pay interest on reserves. But the effect of this change has been obscured (temporarily) under current “liquidity trap” conditions. Because the fed funds rate has hit the zero bound, the interest rate paid on reserves has hit it as well. Therefore, paying interest on reserves is somewhat redundant in this environment, because the rate is at or near zero.

    The change will show up I think at such time as the Fed begins to tighten the Fed funds rate. In a normal quantitative reserve environment, the Fed controls the lower bound for the rate by maintaining only a sliver of excess reserves. Now the Fed will be able to control the lower bound by paying interest on reserves.

    This means that the Fed will not be forced to reverse its quantitative easing in order to implement policy rate tightening.

    This is a world changer, because it will allow the Fed to continue to use its balance sheet as a source of funding for qualitative (credit) easing as long as it desires, while retaining an independent capability to increase the fed funds rate.

    I.e., policy rate tightening is no longer constrained by a requirement for quantitative neutrality.

    In terms of your specific question, this implies in my mind that any program of quantitative or credit easing entered into by the Fed in this credit crisis will be protected from the risk of forced abrupt unwinding, simply due to what previously might have been quantitative reserve constraints in the face of an incipient tightening program for the fed funds rate.

    In other words, the Fed will have more time to unwind current asset programs in the future than what otherwise would be the case. This includes a program of purchasing treasuries.

    In terms of treasury purchases per se, I really haven’t thought a lot about it. But one thing bothers me. My instinct is that, given upcoming financing requirements and relatively depressed bond yields, Treasury should be encouraging market appetite for longer term bond issues such as 30 years. Term extension in funding seems to me like a prudent strategy given the projected size of deficits and attractive costs on a secular basis.

    This would still allow the Fed to buy 10 year treasuries in the market. This might serve the purpose as effective 30 year mortgage duration is closer to 10 years than 30 years. The result would be at the margin a switching of funding from 10 to 30 years with the Fed monetizing at the shorter end and the market buying at the long end. Again, I haven’t thought about this much, but it’s my first instinct for Treasury and Fed strategy.