Robert's Stochastic thoughts

Saturday, May 14, 2011

I tried to comment here but blogger wouldn't let me

My comment.

You get a very respectful link from Krugman, and, thus my congratulations. However, I am even more of a non-monetarist than he is, and I totally disagree with your reading of recent data.

You wrote

a key lesson of recent years is that monetary policy overwhelms fiscal policy. Thus, from 2008-2009 when monetary policy was effectively tight the easing of fiscal policy didn't quite pack much of a punch. Similarly, in late 2010, early 2011 when there was not much fiscal stimulus, but some monetary policy easing under QE2 there was some improvement in the pace of recovery.

Notice how vague the claims are. He didn't type any numbers -- say GDP growth rates or employment growth rates.

I think your perception of macroeconmic events is not related to reality. After the ARRA there was the most rapid accelration of GDP growth over 2 quarters since 1980 and also the most rapid acceleration of GDP growth over 3 quarters. Rapidly declining employment turned into rapidly increasing employment (even ignoring the census).

After the fact estimates of the effects of the ARRA by experts (one of whom chaired the CEA and is biased) are similar to forecasts. Also monetary policy was (by now conventional standards) as loose as it can be, since the now conventional measure is the federal funds rate not growth of monetary aggregates.

Sure the general public considers the ARRA a failure, but that is because they don't compare it to a VAR forecast. The view that it worked as well as expected but was half the right size is absolutely consistent with the data. The public doesn't have the same view of QE2, because they haven't noticed it (frankly it is very hard to notice it).

Your claim about what happened after QE2 is similarly absolutely incomprehensible to me. I note the general view that a shift in monetary policy affects national income and product accounts after "long and variable lags" -- Milton Friedman. The general view is about 6 months after the shift. So standard views on when monetary policy should have an effect place the effect either in 2011 Q1 (very disappointing) or 2011 Q2 (in course). Note that 7 year real interest rates are almost exactly the same as they were when Bernanke first mentioned QE2 and higher than they were after the final announcement of what and how much would be bought.

Also you skipped 2008-9. There was a megahuge expansion of monetary base utterly unprecedented in US monetary history. M1 and M2 massively increased. GDP fell sharply. If one more were necessary, this was the utter final refutation of the quantity theory of money (which I consider to have been utterly finally refuted in the early 80s).

I don't see how anyone can have lived through the past 2 years and think that monetary policy trumps fiscal policy. I would have thought that the utterly overwhelming evidence that points the other way would have convinced everyone (maybe even Friedman himself). I really don't understand how anyone can still be a monetarist.

I didn't mean to be rude, but well whatever, I'll just let it hang out. To me monetarists are like Ptolemaic astronomers (note I have a lot of respect for Ptolemy http://bit.ly/lKt8bK). It fit the data for a while, but now we know that other models fit better.

14 comments:

Noah Smith said...: Robert! You may be right that monetarism is BS. BUT, there are lots of empirical studies out there, some of which are probably a lot better than an informal observation of America's current situation. E.g.:

http://www.esri.go.jp/en/workshop/060914/mihira03_a.pdf

http://www.bis.org/publ/bppdf/bispap19l.pdf

http://ideas.repec.org/p/nbr/nberwo/15565.html

http://www.jcer.or.jp/eng/pdf/kenho2e.pdf

I'm sure there are many more. I assign you to read and critique a bunch of them before concluding that quantitative easing is crap! ;)

I.e., let's actually look through the best available telescopes before deciding that Ptolemy is wrong, shall we? :); 1:56 AM
Noah Smith said...: Robert! You may be right that monetarism is BS. BUT, there are lots of empirical studies out there, some of which are probably a lot better than an informal observation of America's current situation. E.g.:

http://www.esri.go.jp/en/workshop/060914/mihira03_a.pdf

http://www.bis.org/publ/bppdf/bispap19l.pdf

http://ideas.repec.org/p/nbr/nberwo/15565.html

http://www.jcer.or.jp/eng/pdf/kenho2e.pdf

I'm sure there are many more. I assign you to read and critique a bunch of them before concluding that quantitative easing is crap! ;)

I.e., let's actually look through the best available telescopes before deciding that Ptolemy is wrong, shall we? :); 1:56 AM
Mark A. Sadowski said...: You can't comment because no one can comment on your blog.; 7:51 AM
Mark A. Sadowski said...: You wrote:

"Your claim about what happened after QE2 is similarly absolutely incomprehensible to me. I note the general view that a shift in monetary policy affects national income and product accounts after "long and variable lags" -- Milton Friedman. The general view is about 6 months after the shift. So standard views on when monetary policy should have an effect place the effect either in 2011 Q1 (very disappointing) or 2011 Q2 (in course). Note that 7 year real interest rates are almost exactly the same as they were when Bernanke first mentioned QE2 and higher than they were after the final announcement of what and how much would be bought."

Consider the lags. Empirical evidence suggests that peak level effects occur 6 months after implementation. That would be through April at most by my count. The huge boom in job growth occurred December through March. Go figure.

You wrote:
"Also you skipped 2008-9. There was a megahuge expansion of monetary base utterly unprecedented in US monetary history. M1 and M2 massively increased. GDP fell sharply. If one more were necessary, this was the utter final refutation of the quantity theory of money (which I consider to have been utterly finally refuted in the early 80s)."

This is a debate purely in your own mind. Similar runups in monetary base occurred in Japan in the Lost decade and during the Great Depression. And yet empirical evidence still supports the efficacy of monetary policy.

You wrote:
"I don't see how anyone can have lived through the past 2 years and think that monetary policy trumps fiscal policy. I would have thought that the utterly overwhelming evidence that points the other way would have convinced everyone (maybe even Friedman himself). I really don't understand how anyone can still be a monetarist."

After seeing all the evidence I don't understand how you can be a fiscalist.

You wrote:
"I didn't mean to be rude, but well whatever, I'll just let it hang out. To me monetarists are like Ptolemaic astronomers (note I have a lot of respect for Ptolemy http://bit.ly/lKt8bK). It fit the data for a while, but now we know that other models fit better."

Of course you meant to be rude. It's your nature. But believers in Phlogiston Theory are like that:

http://en.wikipedia.org/wiki/Phlogiston_theory; 8:06 AM
Kent Willard said...: The monetary base increased because banks deposited the money from QE back at the Fed as excess reserves.

M2 increased massively? When? It increased pretty much along the long term trend. M1 did jump, as money moved from time to demand deposits - probably due to incredibly low rates, uncertainty, and fear of default.; 7:30 PM
Anonymous said...: M2 didn't go up much in 2008-09. M0 did, the monetary base, "high powered money", the reserves held by member banks in the Federal Reserve went up spectacularily. M0 is the one part of the money supply the Fed can affect at will. And normally, increases in M0 translate quickly into increases in the broader money supply.

BUT this time it didn't. Banks even flush with increased reserves didn't use it to increase lending. They just sat on higher reserves on their balance sheets.

This is the interesting anomaly of 2008-09, how the Feds hugely expanded balance sheet did not translate into an increase in the broader money supply.

Which by the way does not in any way contradict any of your other statements but. . . M1 and M2 did not go up.; 9:21 PM
Robert said...: Kent QE2 starts as deposits at the Fed. The Fed didn't buy 7 year Treasury notes with paper cash.

Anonymous, you wrote "M2 didn't go up much in 2008-09 ... . M1 and M2 did not go up." Get your story straight.

Did M2 go up but not much or didn't it go up at all ? IN Sept, oct Nov and Dec 2008 M2 increased by around 7.5% or around $600 Billion. It went up. It increased at a rate enormously greater than normal. Your second claim is false. It went up (and the increase was much more rapid than the trend increase).

What is the basis for your claim about M1 ? To be mechanical and to dilute the amazing increase, I guess I will look Jan1 2008 through Dec31 2009 (I know you are thinking of a few months but I do this to underplay the dramatic speed of the increase in m1. M1 increased around 24% over that period. This is very much faster than normal. The increase was overwhelmingly concentrated in Sept, Oct Nov and December 2008. Your assertion that M1 did not increase is absolutely false. Yes the increase in base money dwarfed the increase in M1 as it dwarfed everything else in monetary history, but m1 increased rapidly.

I often regret being rude, but I will risk that regret. Check the basic facts before making false claims in the comment section of my blog. Your claim that m1 did not increase proves that you are shockingly ignorant and inexcusably irresponsible.

YOu have made an elementary error. You know that the increase in money was tiny compared to the increase in base money. You decide that it was negligible compared to, say, the second most dramatic increase in money in the past few decades. That something was less immensely huge that the 2008-9 increase in m0 does not mean that it isn't immensely huge.; 10:35 PM
Robert said...: Mark Sadowski

I am replying to your second comment. As you should know, some comment sections are moderated. This is one of them. Comments appear after I read and approve them. I use this feature to screen out comment spam. In practice in recent months (years I think) I have approved all comments (don't tell the spambots).

Your assertion that no one can comment on my blog is rude and reckless. You should know that comments may appear after a delay.

On the other hand, I don't necessarily read comments. I'm free to read or not read as I please. I have not and will not read your longer comment. This isn't because of the short comment to which I am replying. I haven't read anything you wrote in a long time (not since the comment I discussed in a post on Angrybear).; 10:40 PM
Robert said...: Noah

Thanks for the comment.

I concluded that monetarism was dead in 1985 when I read "The Search for a Stable Money Demand Function." Notably I first studied economics in 1985.

The paper hard been written a bit before then. It told the story of how all economists had noted a stable relationship between money and GDP, but then the relationship suddenly vanished. My view was not at all caused by recent events. I consider the proof to be based on data which are now at least 28 years old.

Another way of putting it is that I got my view that monetarism is dead from my first course on macroeconomics. That was what I was taught without any qualification by the professor (Larry Summers).

You will note that I am not the only person who has this view

"When monetarism failed – fighting words, but you know, it really did — it was replaced by the cult of the independent central bank."

Kthug http://nyti.ms/g6MReo

Note that he writes of something *before* the cult of the independent central bank. His view, like mine, must be that monetarism failed in the 80s.

I will stick with the Ptolemy analogy. In 1663 it was not necessary to check the latest observations to be confident that Ptolemy was wrong (brilliant, over a millenium ahead of his time, but wrong).

I utterly reject the following argument, but someone might note that the last defenders of menetarism aren't exactly at top departments (better than my department but not top). It just isn't one of the schools of thought fighting over macroeconomics any more. This may well be because we are in the dark ages (Ptolemy was forgotten for centuries and came back).

My view is the conventional view. The return of the quantity theory of money in the policy debate in the form of quantitative easing is due to desperation. The arguments for QE are not about what the Fed supplies (high powered money) they are about what it buys (currently 7 year treasury notes). It is only called monetary policy, because the Fed is doing it. The Treasury could do something similar if it issued only 3 mont notes and nothing of higher maturity for a year or so.

I will now click your links.

Thanks for the comment.; 10:58 PM
Robert said...: Well I will probably click your links. I have clicked one
http://ideas.repec.org/p/nbr/nberwo/15565.html

and read one abstract. I quote "Quantitative easing translated into a greater and more lasting expansion of M1 relative to nominal GDP." In plain English that means "this episode provides evidence against the quantity theory of money" M1 relative to nominal GDP is the inverse of M1 velocity. The quantity theory is that velocity is stable (not constant) and especially that increased money supply doesn't cause a decrease in velocity. The evidence presented in the working paper does not, by itself, constitute proof that Friedman was wrong wrong wrong, but it points in that direction.

Note the author is most obviously not hostile to QE.; 11:08 PM
Robert said...: Me a third time Noah.

I don't agree with your assertion "some of which are probably a lot better than an informal observation of America's current situation." First I will skip the word "informal"

It is true that examination of recent events in one's native country is not usually the optimal approach to questions in social science, but this time it is (oops qualifier for native born Americans it is -- others should look at the USA too).

There isn't and can't be much other data on the effect of QE on a developed economy in which there short term safe rates are extremely close to zero. IIRC The condition on interest rates leaves current events, the 30s and Japan in the 90s. QE wasn't tried in the 30s. QE has not been tried in countries other than the USA (as far as I know). So there are Japan in the 90s and the current situation in the USA.

The case of Japan in the 90s is not useful, because along with QE there was a very very massive fiscal stimulus. You will note that the japanese public debt to GDP ratio is off all the charts (over 200% always IIRC).

There is, I think, basically no useful information either on QE combined with stable fiscal policy or a shift towards less expansionary fiscal policy. There will be information on one of the two sometime soon (I fear on the second). But only for the USA August 27 2010 and later.

OK now informal. My informal analysis is looking at FRED graphs (it's over at angrybearblog.com). I think that's the right way to do it. There is one event QE2. My view is that the quality of the assets bought in QE is what matters, so I don't consider QE1 to be comparable.

My view is that the argument for QE is crap, because the q stands for quantity and not quality. I think the Fed should buy risky assets not 7 year Treasury notes. My perception that there were good effects of it's 2008-9 purchases of risky assets, along with the absence of the most direct intended effect of QE2 caused this opinion.

The timing of QE2 is controversial (August 27, the final announcement of QE2 in November or the actual purchases). There is no way a model with multiple parameters can be estimated usefully.

I dare guess that my approach is basically very similar to that of the articles you cite (I dare guess that here, because I'm pretty sure no one will read this, more sure than no one will read this and check and, finally, I won't mind much if my daring guess is proven to be totally wrong).

So I am willingly to furthermore daringly guess that my approach is the best approach to evaluating QE possible with currently available data.; 11:42 PM
Noah Smith said...: You're right, it's hard to evaluate policy when your only empirical guide is recent history. The government could afford to pay for large-scale lab experiments to establish the impact of monetary policy in a controlled environment, and it would cost a tiny fraction of one percent of the ARRA. IMHO it would be worth it. But no one is doing this. Doh.

Anyway, just eyeballing the recent historical record, my guess is that QE2 averted a plunge into deflation. Since I buy the idea of Fisherian debt-deflation, I conclude that this saved us from a spike in unemployment. Therefore I conclude very tentatively that QE is a useful policy tool.

But I also guess that QE is not, by itself, capable of restoring an economy to full employment. At the very least, QE of the maximum size that seems politically feasible does not seem capable of restoring full employment.

Even if true, though, this does not mean that unconventional monetary policy is entirely crap. Deflation is scary, especially with high debt levels. Staving off deflation is not nothing.; 11:43 PM
Robert said...: I believe in Fisherian debt deflation too, but the problem with debt deflation is high real interest rates. It doesn't matter if the real value of my debt is ballooning because I pay 3% nominal and prices are falling 3% a year or because I pay 6% nominal and prices are stable.

So for those who are concerned about debt deflation, the key variables are real interest rates. After QE2 the spread nominal-real increased, nominal interest rates increased, from the beginning of actual QE2 open market operations till my post, the real 7 year constant maturity rate increased.

By deciding that QE started when Bernanke first mentioned it (with no details) you can get the effect on the 7 year real rate to almost exactly 0.

You clearly think real rates would have been higher, because there would have been significant deflation. We can agree that we can't know what would have been.

It seems to me that logically all other effects of QE2 should act through the price of the assets that were bought (I can't see how more money in itself could matter and almost no one argued that buying 3 month t-bills would have a significant impact). Why would expected inflation go up when the market price of the assets the Fed was buying went down ?

Given the tiny amount of data, one can argue that the expected change was a price increase (decrease of the 7 year treasury interest rate) but a disturbance disturbed things.

I think we agree that there is very little evidence in the data which are few and with large disturbances. But I think of 2 variables one moved the wrong way and the other did or stayed almost exactly the same depending on arbitrary choices about timing.; 5:36 AM
Noah Smith said...: What about long-term fixed nominal interest-rate debts (e.g. mortgages) that already existed when QE2 went into effect?; 6:35 AM