Scott Sumner  

Williamson pulls rank on Yglesias

The Singaporean Path to Cosmop... Crude Materialism versus the W...

I've often argued that the Fed made a mistake holding rates at 2% at their meeting of September 16, 2008 (right after Lehman failed.) Matt Yglesias has a new piece that is also critical of this decision, and Steve Williamson has a post that is highly critical of the Yglesias piece. However the Williamson piece is riddled with factual and analytical errors.

As an aside, let me just say that I don't get much enjoyment out of hammering the post of a fellow economics blogger, especially one that calls himself a "New Monetarist." But when that individual has commented on my blog, and pointed out his sterling research record, and then asks me where I have published . . . well then the job becomes slightly less unpleasant. And when he says this about Matt Yglesias:

Matt Yglesias's piece is written like a blog, not like a well-researched piece of journalism. He cites the FOMC transcripts, and then provides his own interpretation of what is going on. Apparently, there was no attempt to check this with anyone who might have some expertise. Nevertheless, Yglesias is willing to level the charge that the guy who ran the Fed during the crisis made a "big mistake." Unfortunately, it's Yglesias that is making the big mistake.
Well, let's just say that if you are going to mock people for being unqualified and making mistakes, you really ought to make sure that your own critique is not riddled with errors and shoddy analysis.

Let's start here with Williamson's post:

In particular, reserve balances went from $9 billion on September 10 to $47 billion on September 17, to $104 billion on September 24, to $167 billion on October 1. To put this in perspective, daylight financial transactions through the Fed can be supported with a small quantity of reserves. That quantity was sometimes about $5 billion before the financial crisis. So, $47 billion in reserves is a big number. Indeed, it may be a large enough quantity to drive the risk-free overnight rate to its lower bound, which at the time was zero - the interest rate on reserves. On October 1, 2008, the Fed began paying interest on reserves at 0.25%, and that became the floor rate at that time
Williamson is criticizing Yglesias's claim that Fed policy was insufficiently expansionary, and that the Fed should have cut rates in September. Williamson points to the injection of base money, and the fact that actual market fed funds rates were erratic, and often fell below the 2% target during this period. But Williamson is wrong that the Fed instituted interest on reserves October 1st, or that the rate was 0.25%. The program was actually announced on October 6th, and the rate was set 0.75% below the fed funds target, which became 0.75% on October 8th, when the Fed cut the fed funds target to 1.5%. That's quite different from 0.25%. The Fed's explanation for IOR was correctly characterized by Robert Hall and Susan Woodward as follows:
Oddly, he [Bernanke] explained the new policy of paying 1 percent interest on reserves as a way of elevating short-term rates up to the Fed's target level of 1 percent. This amounts to a confession of the contractionary effect of the reserve interest policy.
Hall has published in some pretty good journals, in case anyone wants to know. Later they said the following:
Raising the reserve interest rate is a contractionary measure. A higher interest rate on reserves makes banks more likely to hold reserves rather than increasing lending. The Fed's decision to raise the reserve rate from zero to 75 basis points just as the economy entered a sharp contraction in activity is utterly inexplicable. Fortunately, the Fed lowered the reserve rate subsequently, but the continuation of a positive reserve rate in today's economy is equally inexplicable. Some economists have proposed that the Fed charge banks for holding reserves, an expansionary policy worth considering.
BTW, I believe I was the first to propose a negative interest rate on reserves in a journal article, albeit not one Williamson reads. It's not a cure-all, but would have helped in 2008.

Stocks fell sharply the day IOR was announced. The announcement didn't initially attract much attention, but for what it's worth stocks also fell sharply over the next couple of trading days. Then the Fed increased the rate of IOR on October 22nd, to 35 basis points below the fed funds target. On November 5th the Fed again raised the IOR, to equality with the fed funds target. Here's Louis Woodhill:

At the time of the Fed's IOR announcement, the S&P 500 was down by a total of 12.18% from its pre-Lehman close, 15 trading days earlier. However, the day that the Fed announced IOR, the S&P 500 fell by 3.85%, and it was down by a total of 17.22% three days later.

On October 22, 2008, the Fed announced that it would increase the interest rate that it paid on reserves. The S&P 500 fell by 6.10% that day, and it was down by a total of 11.11% three days later. On November 5, 2008, the Fed announced another increase in the IOR interest rate. The S&P 500 fell by 5.27% that day, and it was down by a total of 8.60% three days later.

The bigger problem with Williamson's post is that he seems to assume that the primary effect of a fed funds target change is the impact of slightly lower overnight bank rates over the next 6 weeks. Or maybe he thinks that worriers like Yglesias and I think that's the primary impact. But as Michael Woodford showed the main impact of Fed policy announcements is not from the immediate change in interest rates, but rather what it signals about the future expected path of policy. Woodford has also published in some pretty good journals.

The Fed's decision to cut rates 1/4% rather than a 1/2% in December 2007 destroyed one or two trillion dollars in global stock market wealth within a couple hours, but not because markets cared about where the fed funds rate would be over the next six weeks. Indeed markets were so convinced that the Fed blew it that they actually reduced T-bill yields on the news, despite the rate cut being smaller than expected. And the markets were right. Four weeks later Bernanke saw he was far behind the curve (his words) and had an emergency meeting where the fed funds target was cut another 75 basis points, and then another 50 at the official meeting two weeks later. Markets fell sharply in December because they correctly interpreted the Fed announcement as a signal the Fed was behind the curve on the severity of the situation. The markets sensed that the Fed would not be up to the task.

So all the evidence Williamson presents about the Fed's decision not to cut the fed funds target in September 2008 is completely beside the point. It was a signal of the Fed not being sufficiently vigilant against a sharp fall in NGDP, a pessimistic view that turned out correct. A surprising large rate cut on that day (say 100 basis points) would have electrified equity markets worldwide, and at least slightly moderated the severe recession that was already underway. So the Fed did make a big mistake, although the main problem wasn't just the September 2008 meeting, but rather the entire policy regime. There was an unwillingness to be forward looking, to focus on market expectations. Instead the Fed looked in the rear view mirror at past inflation data. There was an unwillingness to do level targeting, to make up for the deflation and falling NGDP of 2008-09. The September decision was a symptom of that flawed policy regime.

Of course I could be wrong. After all, Williamson seems to think that unexpected fed funds target cuts are deflationary, whereas I think they are expansionary. Oddly, asset market prices react to unexpected rate cuts by the Fed as if they are expansionary. But I'd guess that asset market participants don't read the journals where Williamson publishes. So perhaps they don't know enough to move prices in the correct direction.

PS. I'd like to thank Vaidas Urba for pointing me to this post, and helping me to better understand IOR. Vaidas believes Williamson made some other serious errors, such as getting the size of daylight credit completely wrong, but I am not qualified to comment on that claim. Perhaps knowledgeable Fed watchers will add information in the comment section.

Comments and Sharing

CATEGORIES: Monetary Policy

COMMENTS (11 to date)
Vaidas Urba writes:

Thanks for kind words. Some remarks:

- While Williamson is wrong on fed funds rate, I liked his post because of his description of fed funds market: "the actual average interest rate on fed funds trades was substantially different from the effective fed funds rate, as measured". The Fed had lost the control of its instrument. This remains an underappreciated problem, I have always argued that the chaos in the fed funds market did lots of damage, and it was the ECB who implemented innovative policy solutions to solve this problem.

- Switch to IOR regime was expansionary. The expected effective fed funds rate was above the fed funds rate target, IOR regime has eventually solved the problem. My understanding is that stocks fell despite the switch to IOR. While initially the Fed used IOR to lower the effective fed funds rate, after some time the effective fed funds rate dropped below the target. When the Fed raised IOR on October 22 in order to increase the effective fed funds rate and raise it towards the intended target, this was contractionary, and stocks dropped.

- Williamson wrote: "daylight financial transactions through the Fed can be supported with a small quantity of reserves. That quantity was sometimes about $5 billion before the financial crisis."
This is misleading, as peak daylight overdraft was above $100 billion in 2006. In the second half of September - first half of October the quantity of reserves was not sufficient to support payments, as evidenced by the turmoil in fed funds market.

Scott Sumner writes:

Thanks for the info Vaidas. Regarding IOR, in my view any policy that tends to increase the demand for the medium of account will tend to increase the value of the medium of account, and hence be contractionary. That was also the Fed's view, and probably the market view as well. The Fed should cut the IOR by 5 basis points, if only to test that claim.

Vaidas Urba writes:

Scott, assuming that monetary base is fixed, IOR increases demand for reserves. However, this is assumption is wrong, as the Fed was holding fed funds rate target constant. That's why switch to IOR regime was expansionary. IOR reduced the risk of effective fed funds rate overshooting the fed funds rate target, which was a big problem after Lehman.

Shayne Cook writes:


From your post ...

The Fed's decision to cut rates 1/4% rather than a 1/2% in December 2007 destroyed one or two trillion dollars in global stock market wealth within a couple hours ...

Please tell me you didn't actually write that - that it was some sort of editing error or something.

Destroyed wealth ????

Mark W writes:

I enjoyed this but don't look forward to Williamson's response if it's anything similar to his rebuttal the to criticism of his 'raising r is expansionary' thesis.

These comments we're out of themoneyillusion realm I re-enter a world where people question the idea that wealth can be destroyed by poor decisions at the central bank.

Steve Williamson writes:

There is a problem in the blogosphere which I have noticed. A person who actually knows very little (or no more than the average economist perhaps) about the things he or she writes about somehow attracts an audience. This seems to inspire confidence in the blogger's mind and, flush with "success," he/she begins to suffer from the delusion that he/she is in possession of some unique vision for the human race.

Yglesias comes out of that tradition. Wikipedia tells me that he has an undergraduate degree from Harvard, and majored in philosophy. So, he writes a piece for Slate which claims that Ben Bernanke, who has a PhD in economics, a distiguished research record as an economist, and was appointed Chair of the Fed by the President of the United States, made a big mistake. In leveling a charge like that, he had better make a good case. But he doesn't. He doesn't consult anyone who might know something about the decision, but apparently uses his own "expertise" to evaluate the policy decision. So, if "pulling rank" is calling that guy out, I'd say that's a good thing.

In mid-September 2008, the target for the fed funds rate had become irrelevant, so the decision about that target at the September 16 was also irrelevant. You might try to make the case that cutting the target rate might have signaled something, but it's not clear this would have said anything to financial markets that Fed officials were not saying in public or communicating in private. You're correct that I didn't have the details of payment of interest on reserves correct. October 1 was the date (moved forward) when the Fed was permitted to pay interest on reserves. But that's not material to the argument, which is that there was already enough reserves in the system by mid-September that the effective, risk-free overnight rate was zero - the Fed was already at the zero lower bound. Fed funds trading was a mess at that point. We do not have information on the quantities traded, or even prices in a large fraction of the market. And the fed funds market is unsecured, and dominated by risk at that time.

So, learn something about how these markets work. It will help you understand what was going on. Crisis intervention isn't about targeting NGDP.

Hazel Meade writes:

People take Yglesias seriously enough to have to "pull rank"? Really?

I mean hey, Yglesias isn't bad. Every once in a while he says something interesting about price gouging or taxi monopolies, but come on, he's just a blogger.

The first paragraph by Williamson, above, is the logical fallacy of Poisoning the Well, the second is that of Appeal to Authority. Then in the third, he admits he was wrong.

Speaking of needing to learn something.

Vaidas Urba writes:

Brian Madigan has summarized the benefits of switching to IOR regime (FOMC April '08):
"They give us an opportunity to reduce distortions and deadweight losses resulting from our current complex system of reserve requirements. They also give us an opportunity to improve the effectiveness of monetary policy implementation in routine circumstances as well as in conditions of financial stress."

Ryan Murphy writes:

Scott, I think you need to lay it on even more thick. Williamson doesn't seem to understand how ineffectual argument from authority is when he is citing himself as authority.

Yglesias has better economic intuition than most econ PhDs because people like Williamson use math with the veneer of science as a barrier to entry in PhD programs instead of actually teaching economics.

TallDave writes:

Grudging kudos to Yglesias.

Also thanks Scott for again highlighting the oft-overlooked IOR issue.

Comments for this entry have been closed
Return to top