Quantitative Easing
Robert Waldmann
Can the Fed do any more to stimulate the economy ? The question is back. The answer is only by making credible promises about the fairly distant future. My view is that this means no. I review the issue after the jump.
Long ago Paul Krugman proposed that the Bank of Japan target inflation. In the new millennium this is called quantitative easing. The idea is that when an economy is stuck in a liquidity trap, the real interest rate can only be cut by causing expected inflation. The monetary authority can achieve this by promising to cause inflation when it can. It can only after the economy is out of the liquidity trap. So only if the monetary authority can pre-commit to high inflation when things are back to normal and it will want low inflation, can it stimulate investment in an economy in a liquidity trap.
The argument (due to Krugman of course) is that right now the monetary authority can only push on a string. Private agents are willing to hold unlimited amounts of money given the current short term safe interest rate of almost exactly zero. Only when agents want to do something else with their wealth will it be possible to cause inflation by pouring in more money than people are willing to hold at current prices.
The proposal is that the Fed pump huge gigantic amounts of money into the economy (oops it has already) *and* promise not to buy that money back when prices finally start rising. I quote Brad DeLong
Quantitative easing–pouring a whole bunch of cash in the system with the idea of never reversing the money stock expansion could boost spending and employment considerably by creating expectations of inflation and so reducing the spread–but the Federal Reserve is not going there, and regards the idea with horror, shock, and shame.
I note that the first step, pouring a whole bunch of cash, has been done already and that it isn’t needed for the scheme to work. The Fed doing nothing with the idea of pumping a whole bunch of cash in the system when unemployment is back to normal would work just as well.
Brad also has another approach
I do wonder how much good would be done if the FOMC were simply to stand up and announce that they were raising their long-term GDP-deflator inflation target from 2% to 3%. It might do a lot of good. And it is certainly something the Fed could do without cracking its credibility as committed to low inflation.
It might, but it might not. The Fed can’t keep inflation as high as 2%. The announcement is an announcement about policy in the distant future.
OK assume that the Fed has a precommitment technology so it can set future policy now. Should it commit to higher inflation ?
This is a purely academic deeply hypothetical question. I love those. I have two arguments for no.
1) First a major effect of promising inflation in the future is to drive down the dollar. This would help the US economy at the expense of other countries. One of those countries is called China which is investing huge amounts of Yuan in an effort to keep up the dollar. A Fed vs People’s bank of China war over the exchange rate is a terrifying prospect. Basically the Fed might win, the PBC decide to cut it’s losses by shifting to Euros and suddenly US nominal interest rates are very very far from zero.
Recall what Ben Bernanke used to do. He studied the depression. One lesson is that competitive devaluations were a very bad policy for everyone. He must be thinking of that lesson. I’m sure he is trying to maintain exchange rate peace in our time with the PRC.
2) Aside from that, the policy only works if managers and/or shareholders are smart and managers care about the long term interests of shareholders. That is it works if pigs fly.
The policy is to have inflation low to negative until the unemployment rate returns to normal (because there is no way to avoid this) then have high inflation once it returns to normal (by making unemployment lower than the NAIRU). This means that long term bonds will be cheap (have high yield) and that the return from holding them will be lower than the stated yield (their market price won’t rise steadily as maturity approaches but rather will stay the same until the economy exits the liquidity trap).
The assumption is that firms decide how much to invest based on demand and long term real interest rates. This simple formula fits the post wwII data. However, for that period current inflation is a reasonable forecast of future inflation. The proposal is to make a rapidly changing path of expected inflation. The effect of such expectations on investment can’t be determined from available data.
The argument for investing is that the firm can issue say a 10 year bond and invest in real capital and in 10 years, much of the real value of the bond will be inflated away. For this to work, managers have to believe what the Fed says about monetary policy 3 to 10 years from now *and* they have to care. That is not the way in which managers manage (and they admit it). Larry Summers did a survey once. Many CEOs actually responded. The average response was something like
” to decide whether to make an investment we construct an estimate of the effect of the investment on accounting profits equal to sales plus net inventory investment minus spending on labor, materials and excise taxes, then we discount that flow at an annual rate of 30% and see if the number is positive. If the result is positive we invest”. 30% ! They said they had an extreme short term bias. They also care about accounting.
The inflation targeting proposal is that one should invest, even if one only cares about the short term, as one will gain as the value of debt rises less quickly than the calculated yield on debt. That one can write down the debt marking it to market. This does not work. Investors do not want shares of a firm which claims profits because its own debt has lost market value. The firm’s debt loses market value as the risk of bankruptcy increases. In this case, it is losing value (compared to a silly forecast) because of increasing expected inflation. However, the approach to accounting is one to avoid like the plague.
The firms will gain overt the life of the bond, but they won’t be able to tell investors this in the medium future. The policy should cause increased investment but, in the real world, I think it won’t.
I’m not an economist so obviously I don’t fully understand what is going on here, but I am getting a bit confused here about inflationary expectations. Implicit in this talk about manipulating expectations seem to be the following assumptions:
1. Model consistent expectations
2. Infinite horizons
3 Efficient mkts ( in the “prices are right” sense)
First, whose expectations are we talking about? The buyers of long term treasuries? The general public.? The man on the street is going to expect inflation when he sees it or when there is a good deal of media hype about it. Assuming rational expectations seems to be begging the question. How is a mere promise from the Fed going to do the trick?
Second if treasuries are priced based on expectations of short term movements ( rather than PV of expected returns over their lifetime) how can their price tell us anything out medium term inflationary expectations. Won’t their price reflect inflationary expectations only when the average opinion among mkt movers expects the average opinion to expect imminent inflation? Not when inflation is expected sometime in the future?
well you sure seem to understand economics. Yes indeed arguments about manipulaing expectations are all based on guesses about guesses. They have very much lead policy makers astray — for example by convincing them that reducing inflation will be almost painless if one is monomaniacally determined enough.
In fact, long term bond yields are correlated with stated forecasts of inflation made by self described experts.
The idea of manipulating expected inflation depends on this not being 1 for one (at least in a liquidity trap when short term safe rates are zero) so firms are supposed to be eager to sell bonds and invest, because they think that inflation will reduce the value of the payments they have to make to bondholders. It only works if nominal interest rates done’t go up one for one with managers’ inflation expectations.
Anyway, if you think it is all fancy talk about things we can’t possible understand or predict, then I agree with you.
bakho says:
The problem with monetary stimulus is the lack of ability to repay loans. With the recent wealth destruction, loans are more risky, people don’t have enough collateral, etc.
At this point, loaning the money is close to “giving the money” because of the high default rate. The solution is to give the money, but to give it for things that will return on investment in the future or maintain the value of current investment.
So for example, allow people to stay in a house but pay rent. Have government pay workers now to improve energy efficiency and use money saved on energy in the future for future revenues. We need to borrow from the future to reduce opportunity costs today.
The prime obstacles to running a countercyclical policy is the NoNewTaxes idiots that are wrecking our economy and our country.
Dear Bakho
I totally agree. So do Krugman and Delong. Further fiscal stimulus along the lines you describe would be an excellent policy.
However, it seems that congress won’t enact such programs so one has to look for something that can be done without their approval.
In practical terms, I favor soaking the rich (especially bankers) to pay for the programs you mention. The public seems irrationally afraid of deficits. The public has been very eager to soak the rich for decades. A very robust stylized fact about polling is that a majority is in favor of anything which involves higher taxes for the rich.
Such taxes will do little harm right now since spending by the rich is not limited by their after tax income as much as by the lack of time to spend the money.
Of course the rich own congress (and much of the Obama administration) so it won’t happen, but it would be good policy and political dynamite.
The whole committment problem thing seems like it could easily be solved by merely reframing the issue in terms of price level targeting instead of rate targeting.
I was also under the impression that currency devaluation during the depression was widely considered a Good Thing for everyone. If you devalue your currency against mine then I devalue mine against yours then we have both just devalued our currencies against real goods and have increased inflation (a good thing in a recession).
Your comment about agency problems in financial markets is too clever by half. Agency problems abound, but it’s certainly not the case that manager actively wish their shareholders ill. If there is obvious money to be made from the difference between long term interest rates and short term ones, I think it’s hard to make the case that no one will try to make it.
Your point about high required interest rates by CEOs is a reasonable one. I have had some manufacturing experience, and this puzzled me for a long time. However, I now think that oft quoted absurdly high interest rates like 30% are misleading. After investigating this for a while, I asked one cricumspect head project manager that explained that they use absurdly high interest rates (like 30%) to counteract the biases of people who come up with projects and that they actually use much lower rates when it comes time to actually decide on a project. I am still not clear on why they use a high interest rate instead of some discounting factor, but I am also under the impression that businesses frequently do large projects that are considerably under their quoted rate.
I look forward to Sumner’s inevitable response to this post because a lot of the things you said here sound confused, though it’s hard to say for sure because I am certainly not an expert in this field.
Mr. Waldmann,
You wrote:
“Recall what Ben Bernanke used to do. He studied the depression. One lesson is that competitive devaluations were a very bad policy for everyone. He must be thinking of that lesson. I’m sure he is trying to maintain exchange rate peace in our time with the PRC.”
I’m not sure I agree. The principal lesson I get from the Great Depression was that countries’ recoveries from the Great Depression largely followed in the order that they devalued their currencies (see, for example, Barry J. Eichengreen, Golden Fetters). While a sudden devaluation perhaps would be problematic in our day, a faster but orderly devaluation just might be what the doctor ordered.
And I don’t buy the “we need the Chinese to buy our debt” argument. Rather, we need the Chinese to stop undermining our efforts to expand our money supply.
In the final analysis, however, Bernanke may be too timid to do what is necessary, which is quantitative easing on a scale that would drive the dollar hawks into a fit of apoplexy.
What do we need? Inflation!
When do we need it? Now!
Hmm my discussion of agency issues was related specifically to investing in fixed capital. This can be profitable if one is allowed to mark the fixed capital to market but accounts aren’t written that way. The point is the issue is a CEO of a manufacturing firm deciding whether to build a new factory. Traders will try to make money by forecasting future interest rates sure, but I’m talking about NIPA investment in physical capital.
The 30% rate was the average from a survey. It makes sense for principal/agent reasons. It is a social problem. Basically people often say that CEOs have a short term bias and CEOs confirm that claim when asked.
I want inflation too. How ?
There is an important difference between devaluing and floating. Countries devalued, that is offered more of their currency for an ounce of gold. Floating means getting out of the gold market entirely. Devaluing means taking gold away from other countries. There are old papers on the costly struggle for gold between countries still on the gold standard.
The current relevance is that the People’s Bank of China can wreck the world economy if they get angry enough. Bernanke doesn’t even know who is really in charge over there. He has to worry about what they might do.
Thorstein agreea with Mark Sadowski. A round of competitive devaluations helped tremendously during the GD, not hurt. The lesson from both the Great Depression and Japan that is relevant for today is that the Fed must not do too little. Which kinda sucks, b/c the Fed is doing too little. And since 12/31 of last year, the Fed has shrunk it’s balance sheet. That’s US 1929/1937/Japan1990-2009 level of extreme stupidity…
Like I said, those interest rates are likely misleading.
Another way to think about it is this: if capital investment decisions were frequently that shortsighted (lets say because of agency problems just for example) it would mean there was massive obvious inefficiency. If there was massive inefficiency it would be an active area of experimentation and scholarly research (identifying the specific agency problems etc.). The fact that there is not (and I have looked) strongly suggests this is not a massive problem.
one needling thing to say is that while yes the fed has pumped cash in, it holds securities that mature with principal payback on its balance sheet and so it would have to do something, not nothing, to create this cash forever. in fact these instruments generally pay a coupon which is greater than treasuries (more and more they are MBS from FNM, FRE, and FHA) so they eventually will destroy more cash than doing nothing.
for me the frightening datapoint is the massive increase in productivity. correct me if i am wrong here, please (i really want to be corrected if i am wrong). i think this means that capital is being allocated very, very sparsely and only to investments that potentially clear a very high rate of return. it means that a whole lot of somewhat less productive but still productive investments are not being made, that the hurdle is set too high. is this a reasonable reading?
another way to say that the real rate of interest as seen as an expected forward is too high and needs to be brought lower? or am i misunderstanding this?
also, wouldn’t bailing out the banks (actual bailouts, with newly printed money) give us inflation? or fed/govt could make a political deal with bank management to exchange bailout for a temporary significantly higher marginal tax rate on $1M+ earnings.