In a post yesterday commenting on Paul Krugman’s takedown of a silly and ignorant piece of writing about monetary policy by William Cohan, Scott Sumner expressed his annoyance at the level of ignorance displayed people writing for supposedly elite publications like the New York Times which published Cohan’s rant about how it’s time for the Fed to show some spine and stop manipulating interest rates. Scott, ever vigilant, noticed that another elite publication the Financial Times published an equally silly rant by Avinah Persaud exhorting the Fed to show steel and raise rates.
Scott focused on one particular example of silliness about the importance of raising interest rates ASAP notwithstanding the fact that the Fed has failed to meet its 2% inflation target for something like 39 consecutive months:
Yet monetary policy cannot confine itself to reacting to the latest inflation data if it is to promote the wider goals of financial stability and sustainable economic growth. An over-reliance on extremely accommodative monetary policy may be one of the reasons why the world has not escaped from the clutches of a financial crisis that began more than eight years ago.
Scott deftly skewers Persaud with the following comment:
I suppose that’s why the eurozone economy took off after 2011, while the US failed to grow. The ECB avoided our foolish QE policies, and “showed steel” by raising interest rates twice in the spring of 2011. If only we had done the same.
But Scott allowed the following bit of nonsense on Persaud’s part to escape unscathed (I don’t mean to be critical of Scott, there’s only so much nonsense that any single person be expected to hold up to public derision):
The slowdown in the Chinese economy has its roots in decisions made far from Beijing. In the past five years, central banks in all the big advanced economies have embarked on huge quantitative easing programmes, buying financial assets with newly created cash. Because of the effect they have on exchange rates, these policies have a “beggar-thy-neighbour” quality. Growth has been shuffled from place to place — first the US, then Europe and Japan — with one country’s gains coming at the expense of another. This zero-sum game cannot launch a lasting global recovery. China is the latest loser. Last week’s renminbi devaluation brought into focus that since 2010, China’s export-driven economy has laboured under a 25 per cent appreciation of its real effective exchange rate.
The effect of quantitative easing on exchange rates is not the result of foreign-exchange-market intervention; it is the result of increasing the total quantity of base money. Expanding the monetary base reduces the value of the domestic currency unit relative to foreign currencies by raising prices in terms of the domestic currency relative to prices in terms of foreign currencies. There is no beggar-thy-neighbor effect from monetary expansion of this sort. And even if exchange-rate depreciation were achieved by direct intervention in the foreign-exchange markets, the beggar-thy-neighbor effect would be transitory as prices in terms of domestic and foreign currencies would adjust to reflect the altered exchange rate. As I have explained in a number of previous posts on currency manipulation (e.g., here, here, and here) relying on Max Corden’s contributions of 30 years ago on the concept of exchange-rate protection, a “beggar-thy-neighbor” effect is achieved only if there is simultaneous intervention in foreign-exchange markets to reduce the exchange rate of the domestic currency combined with offsetting open-market sales to contract – not expand – the monetary base (or, alternatively, increased reserve requirements to increase the domestic demand to hold the monetary base). So the allegation that quantitative easing has any substantial “beggar-thy-nation” effect is totally without foundation in economic theory. It is just the ignorant repetition of absurd economic prejudices dressed up in high-minded sloganeering about “zero-sum games” and “beggar-thy-neighbor” effects.
And while the real exchange rate of the Chinese yuan may have increased by 25% since 2010, the real exchange rate of the dollar over the same period in which the US was allegedly pursuing a beggar thy nation policy increased by about 12%. The appreciation of the dollar reflects the relative increase in the strength of the US economy over the past 5 years, precisely the opposite of a beggar-thy-neighbor strategy.
And at an intuitive level, it is just absurd to think that China would have been better off if the US, out of a tender solicitude for the welfare of Chinese workers, had foregone monetary expansion, and allowed its domestic economy to stagnate totally. To whom would the Chinese have exported in that case?
Uneasy Money 
David, uncharacteristically narrow-minded.
The fx effect of QE is mostly because people start chasing non-domestic assets. That’s it. And no one wants a capital flight as a result of monetary policy, so the fx effect has to be necessarily sterilised.
So for most export led economies, QE will indeed be a beggar thy neighbour policy, because the non-QE follow on (contractionary open market ‘sales’) is more or less a given. As always, ceteris is not paribus.
It was different when there was gold. Gold set a floor under real interest rates. Then simultaneous QE by diff countries could still be the kind of pure monetary expansion you describe here.
If you want some empirical evidence, just watch out for Korea’s equity markets everytime Japan/BoJ announces something new.
(None of the above is supposed to imply anything about the right course of current policy. Just that the analysis falls short)
LikeLike
“It is just the ignorant repetition of absurd economic prejudices dressed up in high-minded sloganeering about “zero-sum games” and “beggar-thy-neighbor” effects.”
David, would you describe it as “bullshit?”… I mean that in the technical sense of the word of course :^D
LikeLike
When the monetary authority buys long-dated bonds as part of a QE operation, it raises the price of those bonds and depresses the yield. This increases demand for alternative assets pushing down their (expected) yields. One of the ways this manifests itself in relation to overseas assets is through exchange rate depreciation. People react to falling dollar yields by trying to buy euro (say) to acquire euro assets. They stop when the euro starts to look expensive and so euro assets no longer look so attractive.
The extent to which this happens, and the timing, depends on a lot of things. But it has nothing to do with base money. You’d get the same effect if purchase of long dated debt was financed by issue of short dated paper (provided the short rate remained pegged). It’s all just about relative expected rates of return.
LikeLike
Excellent post.
LikeLike
Ritwik, Sorry to disappoint, especially after your many favorable comments in the past.
“The fx effect of QE is mostly because people start chasing non-domestic assets. That’s it. And no one wants a capital flight as a result of monetary policy, so the fx effect has to be necessarily sterilised”
I do not understand this at all. QE would not be QE at all if it were sterilised (actually don’t like that term, but I’ll go with it here). It is exchange-rate intervention which has to sterilised in order to reduce the real exchange rate. If you don’t like Corden’s theory, fine. but all I am doing is applying it.
Please explain to me what you mean by gold setting a floor under real interest rates. Are you simply saying that there could be no inflation under a gold standard, so the zero lower bound became a floor for real as well as nominal interest rates?
Tom, I’ll let you choose whatever definition you like.
Nick, I chose to express the mechanism in the terms of traditional monetary theory. I don’t think the main point really depends on how the causal mechanism is described.
Peter, Thanks.
LikeLike
Nick Edmonds
“When the monetary authority buys long-dated bonds as part of a QE operation, it raises the price of those bonds and depresses the yield.”
“But it has nothing to do with base money.”
I’ve read event studies like Krishnamurthy and Vissing-Jorgensen (2011), and it’s true that statistically significant decreases in (real) bond yields occur on days of US QE *announcements*.
But VAR evidence shows that US QE Granger causes and is significantly correlated with *increased* US bond yields.(And this is precisely what monetary theory predicts it should do.)
And VAR evidence also shows that US QE Granger causes and is significantly correlated with decreased US dollar bilateral RERs. (Also precisely what monetary theory predicts it should do.)
So if empirical evidence matters (and I think it should), I would argue that this has everything to do with base money. Consequently I think that I agree entirely with David’s description of the mechanism.
LikeLike
Guilty as charged, I was too kind. 🙂
LikeLike
Scott, Indeed!
LikeLike
Mark Sadowski,
Your results there are very interesting. I’m intrigued by the lag though.
I certainly think that there are lot of different mechanisms at work and that these may work in opposite directions. So, I have no problem with the idea that an announcement of QE may raise expectations of inflation (whether or not due to the base money implications), which will tend to raise bond yields – countering to some extent the effect that I described. However, as we’re talking about expectations here rather than the implications of actual balances, I would expect it to happen immediately, not with a lag of several months. If that was how the world worked it would seem to offer an opportunity for an easy profit. I think there must be some other explanation here.
LikeLike
Nick Edmonds,
I’m intrigued by the fact that you’re intrigued. Scott Sumner also brought up the “easy profit” problem, and more specifically suggested that this contradicted EMH.
The interesting thing is that there are many papers in the VAR literature on the effects of monetary policy shocks which show similar lagged maximum effects on financial markets. For example there’s Christiano, Eichenbaum and Evans (1996), Edelberg and Marshall (1996), Thorbecke (1997) and Evans and Marshall (1998). These papers in particular all used the same impulse response function approach that I did, namely a recursive strategy (Cholesky decomposition) with the financial market variable ordered last in the vectors. Such an ordering permits an instantaneous response, but it does not prevent the maximum response from occurring with a lag.
I’m much less familiar with the finance literature on EMH, but in my readings I cannot find any reference at all to the fact that so much of the VAR literature on the effects of monetary policy shocks on financial markets seems to be at odds with EMH. I wonder why?
P.S. This reminds me of an old joke:
A finance professor who espouses EMH is walking along the street with a graduate student in economics. The economics student spots a $100 bill lying on the ground and stoops to pick it up. “Don’t bother trying to pick it up,” says the finance professor. “If it was really a $100 bill it wouldn’t be there.”
LikeLike
Mark Sadowski,
Thanks for the reply.
I’m not even sure I’d go as far as invoking EMH. I just base this on my own casual observations (first hand) that in the bond markets, only today’s news really matters and announcements made a few months ago are all but forgotten. I think my theory would be that any response to altered expectations would happen with strictly zero lag. If I failed to find that, but discovered a significant lagged response, I’d probably try and come up with a different theory as to why that was happening.
LikeLike
Mark Sadwoski, what’s your view of the EMH?
LikeLike
Mark, Thanks for the interesting empirical evidence.
Nick, I agree that the expectational effect should be immediate, but I think that there could be reasons why the effect is not immediate, mainly related to the credibility of the announcement and time lags in how quickly the information is processed by traders when the announcement is subject to interpretation.
LikeLike