People are still confused about my views on monetary policy and interest rates. First of all, no one should assume that they understand what I’ve said in the past on these issues. This stuff is very nuanced, very counterintuitive, and I’m not a very talented writer. So let’s try to first see what is actually true, and then think about what I’ve said:
1. Interest rates are not a reliable indicator of monetary policy. I’ve said that 100 times.
2. NGDP growth is a reliable indicator of monetary policy. I’ve said that 100 times.
3. Long term nominal interest rates are strongly correlated with both the growth rate of NGDP, and the level of NGDP relative to trend. I also make this point quite often. So if NGDP is reliable, and long term rates are strongly correlated with NGDP, why are they not reliable? Because other factors also affect long term rates. Trends in global savings/investment, etc. Still, long term rates are often a good indicator of whether money has been tight.
4. Here’s where some of the confusion creeps in. I might say that the high interest rates of the 1970s were caused by easy money, which raised NGDP growth. If bond yields are 15% it’s a good bet than money is easier than if bond yields are 1%. I make that point often. People might assume that means high interest rates always and everywhere indicate easy money. Not so. They usually indicate easy money, because interest rates are correlated with NGDP growth (and levels), but the correlation is far from perfect.
So far I’ve discussed long term trends. A completely different set of issues crops up when we look at instantaneous market responses to monetary policy announcements.
5. I’ve always claimed that easy money surprises can raise or lower long term bond yields. Perhaps if I do a numerical example, this will be easier to see:
Suppose an announcement of a big bond purchase by the central bank would tend to lower bond yields by 20 basis points via the liquidity effect. If there are no other effects, then bond yields will fall. But the policy might also increase expectations of NGDP growth. How much would that factor increase bond yields? Anywhere from 1 basis point to a 1000, or more. Thus the total effect could be rates falling 19 basis point, or rising 980 points. Nominal rates would likely fall if the bond purchases are viewed as an ineffective gesture by the central bank, which will not be pursued aggressively enough to dramatically change the expected path of NGDP. I.e. if it’s not likely to raise rates at least 20 basis points because of faster NGDP growth. Bond yields would rise if the action is expected to lead to 1979-81 type double-digit inflation. We all know how high T-bond yields got in the late 1970s. If the Fed engineered very fast NGDP growth expectations, it could happen again.
6. Even worse, when at the zero bound things get even more complicated. The Wicksellian equilibrium rate may be well below zero in Japan, and hence modestly higher NGDP growth expectations may simply push the number closer to zero, and hence nominal rates would not increase.
So there is a frustrating level of ambiguity in this picture, but I’m afraid that’s the way the world is. Neither I nor the Japanese bond market has a very good fix on how determined the BOJ is to raise NGDP growth. There is likely to be some confidence on the size of the liquidity effect, and massive uncertainty on the income and inflation effects. And given that radical uncertainty, you’d expect bond yields to be quite volatile, reacting to hints of future policy intentions coming out of the government. And that seems to be what is happening. Bond yields first fell sharply on the expansionary move, and then rose even higher, presumably on expectations of stronger NGDP growth. I don’t know what news they are looking at, but it wouldn’t take much when there is so much uncertainty. Think bitcoin.
7. In this blog I tend to emphasize cases where interest rates have gone the “wrong way” in response to monetary policy, because I am trying to get people to see things in a different way—shake up their undeserved complacency that easy money always means low rates. But I have never, ever, argued that bond markets always react one way to policy announcements. I emphasize the perverse reactions (Jan. 2001, Sept. 2007, Dec. 2007) in order to show that rates don’t always fall with easy money announcements, not to claim that rates always rise following easy money announcements. That would be absurd.
1. Over long periods of time long term bond yields do tend to track GDP growth (and levels) pretty well. So I’m likely to have made some generalizations in that area equating low rates and tight money. Japan still has tight money! They have low expected NGDP growth. And they still have low rates.
2. As far as the immediate market reaction to monetary announcements, I’ve always argued that it is highly unpredictable, but that there are certain principles that seem useful. An announcement likely to dramatically change the future path of policy is more likely to lead to a “perverse” reaction in bond yields, than would a one-time injection of money. I wish I could say more, but I’m often just as confused as you are.
PS. I try to make very precise statements when I write posts, but often I fail to be precise enough. So it’s partly my fault. But some burden also falls on the reader. I have readers who think I am “pro-China” based on the “feeling” they get from reading my China posts, even though I am strongly opposed to government policies there, and say so. I’m sure some of my posts left the “feeling” that an expansionary monetary policy surprise will always raise bond yields. But I’m quite certain that I’ve never made that claim.
PPS. James Alexander sent me a very good article by Ben Southwood in the British publication “CityAM.” I’m told that “City” workers often read it on the train while going to work.
PPPS. I’ve recently had lots of problems with my blog. It crashes quite often. I don’t know why. My tech support hasn’t been able to figure out a way to limit the lengths of comments, so Geoff is still free to ramble on interminably. My spell check no longer works–it says “no spelling errors” when the post is full of errors (spelling errors, not content!!) So you may notice even worse spelling than usual. Feel free to send in any suggestions, I use WordPress.
PPPPS. My prediction on the next shoe to drop—-the ECB!