Monday, October 25, 2010

Fed governor: If we could guarantee 5% NGDP growth, it would be great

Mr. Dudley, President of the NY Federal Reserve Bank, is speaking at Cornell today. One of the economics faculty got him to make a presentation in his class. He could speak a little more freely in exchange for no audio or video recording. I eagerly attended and took copious notes. Though I will eventually put out most of them, I wanted to jot down a few of the more important points now. I have tried to be accurate, but recognize my interpretation of what he said may not be direct Fed policy. He will have another, public lecture later on today that I'm sure will be more widely published.

His comments were very forthright and illuminating about how the Fed views the current recession and the Fed's role going forward. In starting out, he said that the Fed's first tool is clear communication of commitments, and then he talked about clarity. This shocked me. Clarity has not been the Fed's best tool in recent decades. He later mentioned that one of the Fed's great failures had been a lack of clarity during the bailout phase. They figured that if the end came out alright, people would just pat them on the back and say "good job." But they were making some ugly choices, some unfair choices, and didn't make enough effort to explain them. So stories started circulating about what the Fed was after that weren't correct, and that hurt Fed credibility. Now they are improving clarity. Dudley led by example.

The second tool when interest rates are 0 is QE2 (the second round of qualitative easing). He drew a nice little graph (like this one!) for everyone with marginally decreasing benefits to quantitative easing and marginally increasing costs. The Fed is trying to do two things: 1) Figure out where the marginal benefits to quantitative easing is equal to the costs and 2) Lower the cost curve. "And I am here today to affect the cost line and shift it down." The more people believe the Fed's commitment to exit, to mop up excess reserves, to fight inflation when it comes back, the lower the cost curve will be and the more easing the Fed can do now.

[Back to me:] Paying interests on reserves (IOR) has been fingered by Sumner, Beckworth, and others as a cause of much of the monetary contraction since 2008. Yes, the Fed sent out $800 billion, but all of it went straight back into Fed deposits instead of into the economy. It shored up banks and kept more of them from failing, but didn't produce growth. The big take away for me was understanding what the Fed thinks it's doing, keeping up a contractionary policy while claiming it's trying to ease more.

[Back to Dudley:] The answer is that IOR is a major new tool to reduce the costs of quantitative easing. It allows the Fed to convince investors and other people that they can and will mop up excess reserves later. By shifting the cost curve down, it allows them to do more easing. "We can control the demand for credit" by changing the costs of credit directly. The combination of reserves which cost them 25 basis points (0.25%) and the long term assets they purchase at about 4% return means that the Fed is currently bringing in an $80 billion annual profit, so they are very popular on the Hill right now. If, in order to mop up excess liquidity, they have to raise the IOR to more than 4%, however, they start losing money. So the risk of QE2 is that it becomes a major liability later.

Someone asked what the exit strategy was. His response was that it's not the concern right now. Right now we need to get unemployment down. There is a disagreement on the committee which order to do things when they exit (divest themselves of what they purchase during QE or raise IOR), but it will depend on the situation at the time and whether they need to raise short term or long term rates first. He talked about several other tools at the Fed's disposal, like reverse repurchase agreements and allowing banks to put reserves in less-liquid assets. The key quote: "Inflation is too low right now" for them to worry about exit.

[Back to me:] When is the time to exit? At one point, he tossed a couple numbers into the air. Until aggregate demand has grown enough that unemployment is back down to below 8% or if inflation got above 2%, they really aren't looking at exit.

I had communicated with Sumner about Dudley's arrival and asked what question I should ask if I got the chance. I got the chance. Given that the Fed's purpose right now is to lower the cost curve, it seems to me that there is a political economy question of how to sell it. If stimulating aggregate demand is your goal, why not tell people you are targeting a proxy for aggregate demand, like NGDP growth, and announce you are working on raising people's average incomes rather than trying to raise inflation? His response:

[Dudley:] That is the right point. We would like to see something like 1.75% inflation. It's too low right now. We are worried about NGDP because people can't leverage. [He made several statements about why low NGDP growth is bad.] We could try targeting NGDP, but would it be credible? Could we actually hit our targets? There's also a communication problem. People hear "nominal income" and don't know what to do with it. But this is a key issue. If we could guarantee 5% NGDP growth per year for several years, it would be great.

[Me:] Personally, I score a big win here for Sumner et al. I'll do another post tomorrow about his comments on the recession itself and the steps Congress, the Fed, and international policy makers are doing to prevent this all from happening again.


  1. Good on you for posting the Q&A. Hope your blog gets linked in the FT Alphaville blog and elsewhere.

  2. Good job, my dear Watson.

    What is the real expectation that we can guarentee any growth?