The Taper Tantrum

I am marking my beliefs to market as Brad DeLong does from time to time. One of my old beliefs whose market quotation has declined is extreme skepticism about the effectivenss of US monetary policy at the zero lower bound.

This skepticism was largely caused by what I perceived to be the weak effects of the second round of US quantitative easing (QE2) and what I perceived to be much weaker arguments that it had been highly effective based, in my view, on the use of extremely convenient and highly variable lags. Also interest rate shifts of two or three tenths of a percent were treated as extremely important. The ability to reliably predict such shifts would be worth a huge amount to bond traders, but macroeconomic research focused on the effects of interest rates gives no hint that they would have a reliably detectable effect on output, employment or inflation.

[I am moving some borderline relevant text after the jump. The bottom line is that Abenomics shows nonstandard monetary policy can have the effects which I doubted it could. I think the open debate is about what the FOMC can do given the federal reserve act and the political environment.]

A dramatic event, the taper tantrum, strongly suggested that my view of the effects of nonstandard monetary policy in the USA was totally wrong. As I vaguely recall, the taper tantrum was a dramatic increase in long term bond rates and a dramatic decline in inflatin breakevens which followed immediately after a FOMC press conference in which the tapering off of QE3 was discussed. Googling back, I see it was much more complicated than that.

First a sincere attempt to time the taper tantrum gave me the date May 22 2013.

Exactly one year ago today, on May 22, 2013, the Federal Reserve announced that it would begin tapering back its roughly $70 billion a month in bond and mortgage backed securities program. The news made risk assets the ugly duckling of financial markets.

FRED caused me to ask “What tantrum ?”


Nothing much happened around the data of Bernanke’s testimony.

Further googling revealed a four week gap between the first public extremely official discussion of tapering and the tantrum

positive economic news in the spring of 2013 led Federal Reserve Chairman Ben Bernanke to testify to Congress on May 22, 2013, that the Fed would likely start slowing—that is, tapering—the pace of its bond purchases later in the year, conditional on continuing good economic news. This testimony laid the groundwork for the June 19 press conference in which the Chairman optimistically described economic conditions and again suggested that asset purchases might be reduced later in 2013: “[T]he Committee currently anticipates that it would be appropriate to moderate the monthly pace of purchases later this year.”

the tantrum occured after the press conference whose content seems to add little to Bernanke’s testimony. Notably less than a year later, a Forbes contributor Kenneth Rapoza seems to have conflated the two announcements. We seem to be back to convenient and variable lags.

But even the June events do not fit the story as I vaguely recalled it. There was a market response whose timing matched the press conference, but it consisted of an interest rate increase of about 30 basis points and a breakeven decline of about 20 (meaning real interest rates increased about 10). There is no way that such changes can be a big macroeoncomic deal. Again events which matter a great dea to bond traders were discussed as if they have important effects on people in general.

Zooming out, interest rate shifts in the Spring of 2013 are large enough to potentially make a noticeable difference — long term rates including mortgage rates went up by about 1%. But the identification of the whole shift with the taper announcement is based on stressing timing when it is convenient to do so and ignoring it when it is inconvenient. Either the few days around June 19 2013 are the topic (and there was no similar news May 22) so interest rate shifts are too small to be worth all the fuss or properly called a tantrum, or a long interval of time matters so the claim that the issue was the taper press conference (and not of course the taper testimony) is based on ignoring the event studies method not on using it.

update: graph and some discussion added


Notice that the small quick late June shifts have the pattern one would expect given monetary tightening — a nominal interest rate went up, expected inflation went down, the S%P 500 index declined and the dollar appreciated against the Euro. However, the large slow shifts from the end of April through early September do not show that pattern at all. The nominal interest rate went up, expected inflation went down, the S&P 500 went up, and the dollar depreciated (slightly). Higher nominal interest rates and stock prices correspond to good news about growth. The overall pattern seems to me to fit a decline in the expected price of petroleum. But in any case, it doesn’t look at all like the effects of tighter monetary policy or tighter expected future monetary policy.

Totally aside from timing (which is critical for identification except when it is inconvenient and irrelevant) the pattern suggests that most of the shift in Spring and Summer 2013 was not a tantrum due to the taper. The dramatic shift June 19th got attention. The large shifts over the whole period looked important, and, given vague memories, seem to have occured about the time of the taper (at least that is what I carelessly thought when looking at long time series). But they clearly are different — occuring over different intervals and with opposite sign changes in Stock prices and the dollar Euro exchange rate. The whole graph doesn’t support the idea that anything the FOMC said or did had a large enough impact on the economy to be noticeable. In a way, this is good news for those convinced that US monetary policy at the ZLB matters. US economic growth accelerated about a quaerter or two after the taper tantrum. Anyone silly enough to think a 30 basis point shift in interest rates should cause a major adjustment to expected future GDP growth would have gotten the sign of the change in growth rates wrong.

end update:

In conclusion, I think the standard analysis of the taper tantrum (including mine until just now) is of the same quality as the analysis of QE2 which drove me (temporarily) nuts. A story fits the data if magnitudes are ignored and if exact timing is both critical and irrelevant.

Appendix: stuff I moved to after the jump. I forgot that you have to click [Read More] to see the graph.

I recall that I was actually enthusiastic about QE3 (based on open ended purchases of mortgage backed securities rather than 6 year Treasury notes). I was sincerely disappointed by the limited evidence of much in the way of effects on mortgage interest rates. I became very convinced that US unconventional monetary policy had limited effects in December 2012, when markets responded almost not at all to the Fed’s announcement that it was going to follow an Evans rule (also called QE infinity and by me only QE 3.1). This is the belief which I will mark to market.

The same month Shinzo Abe and the Liberal Democratic Party (a terrible name for a rock band) won a solid majority in the Japanese Diet and announced a radical reflation program including quantitative easing. Abe appointed ultra dove Bank of Japan critic Haruhiko Kuroda governor of the Bank of Japan (google warning he shares a family name and first initial with some baseball player). Kuroda said he would do anything necessary to achieve 2% inflation and announced a huge QE. The dramatic effect on the exchange rate, stock market indices and the difference between returns on nominal and inflation protected bonds (inflation breakevens) convinced me that unconventional monetary policy can be effective. Weaseling, I note that I had argued that the US problem was that the Fed Open Market Committee couldn’t signal policy to be implemented after the terms of current committee members ended. Unweaseling, I admit I was skeptical about Abenomics in general and inclined to argue that the roughly simultaneously announced fiscal stimulus was key. I have since admitted that Abe and Kuroda showed that inflation expectations can be changed by a radical shift in monetary policy can be achieved, and fallen back to the claim that extraordinary events such as a radically new solid parliamentary majority and a very determined head of government are required (because independent central banks are independent because of laws which can be changed).

In any case, Japanese inflation expectations appear to have been successfully managed and to have caused higher output (including construction) as should be the result of the resulting reduced expected real interest rates. It is important to note that the extremely radical expansionary monetary policy was not enough to prevent a recession starting Spring 2014 following a 3% increase in the value added tax. Monetary policy at the ZLB isn’t helpless, but it can be overwhelmed by fiscal policy. The assertion that a sufficiently determined monetary authority can target nominal GDP has been pretty much disproven (again).