Concrete steps for Mr. Cochrane
By Scott Sumner
John Cochrane has a post on Japan’s long period of near-zero CPI inflation, and then adds this footnote:
Update David Beckworth on the same topic. I’m less of a NGDP target fan. It’s like saying all the Chicago Cubs need is a “win the world series” target. OK, but what do you want them actually to do differently? What 3 trillion of QE wasn’t enough, but 6 will do the trick? I know the answer, that talk alone tweaks some off equilibrium paths to generate more “demand” today. And monetary policy does seem to be just talk these days. But still… I’m also less of a fan of looking at monetary aggregates. At zero rates, money = bonds, and MV=PY becomes V = PY/M. But it’s a well stated analysis in these terms, and nice coverage of the fiscal theory at the end.
I must say I don’t like this footnote. The off hand remark about the Equation of Exchange doesn’t really make any sense. John probably has in mind something about the quantity theory breaking down at zero-rates because V becomes unstable, but my pet peeve is when people confuse the Quantity Theory (which is a theory) with the Equation of Exchange (which is an identity.)
The more important issue is John’s request for specific actions that the Japanese could take to hit their NGDP target. Unfortunately, Prime Minister Abe announced a
6 600 trillion yen NGDP target without mentioning a date (which sort of defeats the whole purpose.) But I’m going to go with David Beckworth’s estimate of 2020 as being roughly the target date. That’s the sort of time period that makes sense for an elected politician. A 50-year target would be just silly, as Abe won’t be around then. Also note that Japan’s NGDP is currently 5 500 trillion yen, so it’s easier to think of the target as a 20% increase over 5 years, which is slightly under 4%/year (due to compounding.)
John seems to assume that market monetarists like David and I think in terms of money supply targeting. But that’s old school monetarists. We favor a policy stance that the market expects to lead to on target NGDP growth. There are many ways of doing this, and not surprisingly all have drawbacks. But we live in a messy imperfect world, so what do you expect?
1. NGDP futures targeting. This is the optimal policy, in my view. Have the BOJ peg the price of a 5 year forward NGDP futures contract, in order to equate the policy target with the market forecast.
Problems include a possible risk premium, and the need to set up such a market. The latter problem is easy to solve in a technical sense, but perhaps not a political sense. The risk premium is real, but very unlikely to be of macroeconomic significance. It doesn’t much matter if the actual growth rate is 3% or 5%, rather than 4%. You just want to avoid the negative NGDP growth of 1993-2012.
2. Let’s say NGDP futures are off the table. What other concrete steps are possible? I’d say currency depreciation is a plausible option—it worked for FDR. But how much should the yen depreciate? This is tricky, but becomes much easier if you assume the monetary authority does not control the real exchange rate in the long run, and that 2020 is far enough out to be the long run. Then you estimate the current market expectation of Japanese NGDP growth between now and 2020. It’s probably about 5% total, or 0% RGDP and 1% inflation per year over the next 5 years. So we are currently about 15% short of the 20% target in terms of market expectations.
Now consider that the current value of the yen reflects those same market expectations. So the BOJ should do a 15% depreciation of the yen against a trade weighted basket of exchange rates, and then peg it at that new level. If the 5-year forward price is different from the spot price, you’d probably want to depreciate the 5-year forward exchange rate by 15%. There are some technical issues here relating to how quickly you’d like the policy stance to converge to the new target.
The drawback is that other countries would complain. Let them squeal! It’s time for the Japanese to do something about deflation. In any case, ANY effective policy will cause the yen to depreciate, as we saw after Abenomics was announced.
3. There are many other market approaches, which all have drawbacks. You could instruct the research staff of the BOJ to construct a model linking NGDP expectations with a basket of market indicators, such as TIPS spreads, equity prices, exchange rates, etc. Then instruct the BOJ to do open market operations until this basket of indicators predicted on-target NGDP growth. You could even do an FDR, and depreciate the yen by 15% against gold. But that’s risky if gold prices are volatile (and they are).
You may wonder how these solutions solve the so-called “liquidity trap” problem. They don’t, but they do make it easier to see that there actually is no liquidity trap problem; it’s a myth. To his credit, Paul Krugman has been very upfront in acknowledging and even defending the implications of the liquidity trap model. If the central bank can’t create inflation, then ipso facto it can’t influence the nominal exchange rate—at all. But we know from recent history in Japan and Switzerland that central banks can influence the exchange rate quite dramatically, even at the zero bound.
Interestingly, there was the same sort of skepticism in the 1930s. Naysayers claimed FDR would not be able to devalue the dollar against gold, via his gold buying program. They claimed that he’d be able to influence the Treasury’s buying price of gold, but not the international market price. But they were proved wrong. He raised the official price from $20.67 to $35 an ounce, and made it stick in the international markets. So in fact there never was a liquidity trap problem, but this needs to be rediscovered over and over again by economists who are not familiar with monetary history. One example of the rediscovery of FDR’s policy was Lars Svensson’s proposal “foolproof plan” for exiting a liquidity trap.