Ramin Toloui on QE

David Beckworth directed me to a new paper by Ramin Toloui, which examines the various transmission mechanisms for quantitative easing. This might be the single best paper I’ve read on the effects of Fed policy during the Great Recession and its aftermath.

In the past, I’ve often criticized people who judge the effectiveness of QE by looking at the impact on interest rates. Toloui does a nice job explaining why macroeconomists should avoid reasoning from a price change:

First, evaluating the effectiveness of quantitative easing by focusing solely on realized bond yields is inherently flawed. The conventional narrative for QE’s transmission mechanism to the broader economy focuses on the causal chain whereby central bank purchases boost bond prices, thus stimulating investment and consumption via the intermediate channel of lower bond yields. But this ignores the powerful expectations channel through which central bank policy operates. If central bank actions positively shock expectations for future economic prospects, that may directly shift the willingness of businesses and households to invest and consume. Success in breaking deflationary expectations can catalyze increased consumption and investment.

But such a shift toward reflationary expectations—higher growth, higher inflation—also tends to increase bond yields! In theory, therefore, the impact of “successful” central bank balance sheet policy on realized risk-free yields is ambiguous. At minimum, any central bank success in generating reflationary expectations would mitigate observed downward effects on yields, understate the full impact of QE policies, and help account for why bond yields increased during some implementation periods cited by QE skeptics.

Then Toloui develops a very clever way to address the identification problem:

First, this study explicitly models how changes in Fed balance sheet policy shifted market expectations for the Fed’s future policy reaction function. It achieves this by using a Taylor Rule-based framework to model the market’s future expected policy rate as a function of the market’s expectations for future inflation and output. Controlling for these variables makes it possible to identify periods when the market perceives there to be outright shifts in the central bank’s future policy reaction function—that is, when the market expects that the central bank will select a different policy rate in the future than it would have previously given the identical expectations for inflation and output.

It is useful to illustrate this concept with an example. Assume that the month before a Federal Open Market Committee (FOMC) meeting, the market expects that the policy rate three years in the future will be 2 percent, expected inflation will be 1½ percent, and the expected output gap will be zero. The FOMC then meets, leaving the policy rate unchanged but taking other actions. In the month after the FOMC meeting, the market shifts its expected policy rate to 1¼ percent while the market’s expectation for inflation and the output gap remain unchanged at 1½ percent and zero, respectively. This indicates that the market anticipates that there has been an accommodative shift in the central bank’s future policy reaction function—i.e., the market expects the Fed to choose a policy rate that is ¾ percentage points lower than before, despite identical inflation and output gap expectations.

This is indirectly related to a point I’ve made about “forecast targeting”.  The point is not to get the market forecast of inflation, or the market forecast of future levels of the fed funds rate; rather you want the market forecast of the fed funds rate that is likely to lead to on-target inflation.

Toloui also understands that any evaluation of interest rates should be conditional on the state of the economy.  Instead of using this approach to identify the optimal policy rate, he uses it to identify the impact of monetary policy.  He finds that the effect of QE on interest rates is even greater than estimated in previous studies, when conditioned on the state of the economy.

He also has some very interesting things to say about the impact of Fed policy on riskier assets:

But that is where the future Fed reaction function becomes important. To the extent that the market believes that the Fed will be more quiescent in face of future inflationary pressures, the risk that the punch bowl will be removed diminishes and prospects for a boisterous party increase. The market’s expectation for the Fed’s future policy reaction thus affects the attractiveness of a wide range of financial assets, not just U.S. Treasury bonds.

It is worth emphasizing this point, since investment committees around the world were focusing intensively on Federal Reserve policy during the post-crisis period. The interest of investors in new policy announcements was not limited to the technical dimensions of particular initiative. Rather, market participants were looking for what these announcements revealed about the character of the central bank. Was the Federal Reserve willing to do “whatever it takes” to secure a recovery and clip the tail risk of a deflation scenario? If so, portfolio managers would have a “green light” for investing in riskier assets that would benefit from a reflationary scenario. Each additional policy innovation provided more information about the Fed’s attitudes.

Lots of people have in mind a model where monetary stimulus inflates asset prices by lowering interest rates.  But that’s not actually what’s going on.  Asset prices plunged during 2008-09 even as interest rates were cut to zero.

Toloui has a much better explanation.  Monetary stimulus creates a macroeconomic environment (basically “prosperity”) where risk assets do well.  It has nothing to do with creating bubbles; it’s all about the fact that what’s good for America (in a macroeconomic sense) is more often than not also good for the stock market and credit spreads.

I strongly encourage younger academics to take a look at this paper, it provides lots of ideas that point the way toward future research opportunities.

Tags:

This entry was posted on November 20th, 2019 and is filed under Uncategorized. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response or Trackback from your own site.