What more can the Fed do ?
This post is long, vague, sloppy and after the jump. The one sentence version is that the Fed can affect the real economy by buying assets which private investors consider risky.
Before discussing useful things which I think the Fed can do to stimulate the economy, I will explain at length why I think that some proposed interventions would not be effective. I won’t argue this (here or in comments) but I think that last years massive purchases of treasury securities (aka QE2) had very little effect on anything.
One possible future policy would be much more of the same, that is much larger purchases of long term treasuries with Fed liabilities. The academic discussion of possible QE2 started with a proposal of $ 5 trillion new Fed liabilities issued. The actual QE2 as announced was one tenth that much. I don’t think that much huger operations would have a very large affect on anything that matters. I think this for two reasons.
First the return on long term bonds with negligible default risk is the compounded expected return on short term bonds plus a risk premium. Here the risk is not default risk but the risk of mark to market losses over a short horizon due to higher than expected future short term rates. A change in the amount that private investors must hold reduces the risk premium. The private sector as a whole can go short Treasuries (only the Treasury can do that) so the risk premium can’t be driven below zero. It appears to be very low right now, since the overall interest rate yield on Treasuries is low and presumably short term rates will return to normal some time in the next 10 years.
Second the interest rate that matters for private sector fixed capital investment is the cost of capital to firms, not the cost to the Treasury. Here I think an important problem might be the cost of capital to firms with poor bond ratings, but the risk premium they pay is mostly due to default risk and not at all the same as the risk in long term Treasuries (I’d guess it is negatively correlated — good GDP growth means higher interest rates on short term Treasuries and low risk of default by firms — long term Treasuries are insurance against poor growth — less of them in private hands should, in my view, increase the cost of capital to firms at risk of bankruptcy).
Third, while firms care about interest rates over years, I suspect that managers have enough of a short term bias that the rate that matters is over 5 years or less (in a pinch 7 years or less). The expected flow of funds seven to thirty years from now is just not likely to influence their decisions. These interest rates are already very low and can’t be negative.
I don’t think the Fed can drive down short term nominal rates expected for many years from now with QE3 or announcements about plans for the future or in any way at all. They depend on future monetary policy by future Fed Open Market Committees facing future conditions.
The interest rate that matters for firms deciding on investment should be the real interest rate. I also don’t think that the Fed can drive up inflation expected for many years from now for the same reason. A higher declared inflation target should only affect inflation expected for the near future if the Fed will have some way of affecting inflation in the near future. It shouldn’t affect inflation expected for when the economy is back to normal — a promise to create inflation in the distant future by driving unemployment below the NAIRU is not credible.
So I don’t think that much more of the same will affect GDP much.
How about helicopter drops of money ? If the Fed gave away money rather than loaning it or buying financial instruments, then it would create the illusion of wealth. However, the Fed can’t do this. The helicopter drop story is a teaching example. Giving money away is fiscal policy. This isn’t just a semantic point — Congress can’t give any other body the authority to give away US Federal money — this is clearly written in the constitution. Furthermore the argument that the Fed should give away money is based on the argument that Congress should but won’t. Doing something that only Congress may do, because Congress won’t do it, is plainly unconstitutional. Also I’d guess someone (Ron Paul or a Paulican) would sue. The Republican party is against this, so I guess that the Supreme court would block it.
I think it is already clear what I think the Fed can do which would be useful. First I think the economy would be stimulated if someone bought a whole lot of junk bonds driving up their price and driving down the cost of capital to firms with poor bond ratings. The creation of Maiden Lane, Maiden Lane II and Maiden Lane III ltds makes it clear that the Fed thinks it can’t do this directly. So I propose loaning more money to one of them (or setting up Maiden Lane IV) and having that entity buy a lot of junk bonds.
These are the assets whose price is most likely to affect investment. They are assets which are considered risky so their return can be much reduced if private investors are required to hold less of them.
Other risky assets are, of course, mortgage based securities. The Fed can and has purchoased rmbs issued by government agencies. Here I tend to guess that the Fed can renegotiate servicing contracts to make foreclosure costly to loan servicers. I even think the Fed can renegotiate mortgages with debtors (OK I hope it would be allowed to do this if it tried). In any case, it can definitely agree with firms servicing its mortgages that they are not punished for forgiving debt which will never be paid or rewarded for foreclosing on houses which they then can’t sell.
As beneficial owner, the Fed is the principal. It must be able to prevent its agents from acting against its interests.
So I think that the Fed can act as owner of mortgage backed securities to renegotiate absurd incentive contracts and can (indirectly) buy junk bonds.
So why isn’t anyone talking about these possibilities ?
The Fed buys non investment grade securities and it buys (more) MBS and then restructures those mortgages to provide princpal and interest debt relief. That’s the plan?
This is not the role of the Fed. Where would the Fed get the money to do all these nice things? Print it?
Some things were done in 2008. We were in a crisis then. We are not in a crisis now. Want to destroy this country? Follow these recomendations.
I thought it was obvious that the Fed issue new liabilities (that is print money) to buy more rmbs and non investment grade corporate bonds. Notably the Fed has tripled its liabilities and inflation decline from the targeted 2% to below target. I see no problem with the Fed’s printing a lot more money right now.
I think that a diversified portfolio of non investment grade corporate bonds is an excellent investment. I believe that such a portfolio has outperformed t-bills over every period of a year or more, ever.
Moody’s notes that default rates on speculative bonds are similar to interest differentials compared to Treasuries. Importantly, Moody’s defines default broadly. This would imply similar returns on treasuries and on a diversified portfolio of speculative bonds only if recovery ratios were always zero. One interest payment one day late is counted by Moody’s as a default. Even Lehman senior unsecured debt traded at a positive price, 14 cents on the dollar, after Lehman went bankrupt. The strategy of going long speculative grade bonds (a lot of them chosen at random) and selling *after* default on the secondary market (for well over zero cents on the dollar on average) would, I think, clearly dominate investing in treasuries.
http://www.moodyskmv.com/research/whitepaper/52453.pdf
The only problem with such an investment is that it would increase the Fed’s currently huge profits (the highest profits ever earned by a US based entity) and reduce the Federal budget deficit.
Given the extreme profitablity of the emergency measures of 2008, I think it is past time to make public investment in risky securities normal practice. Of course it could end up with disaster as in the case of Singapore (which was poor when they started doing it). But I doubt it.
Demand Inflation Now!
In the abstract, purchase of lower-quality assets may be a good idea. In reality, the Fed did that, and there was a big internal backlash. Bagehot is still the granddaddy of central bank self-restraint, and he said cental banks should only by high-quality assets.
The suggestion to have the Fed purchase low-quality assets has been around for some time, and would serve the purpose of narrowing spreads for low-quality borrowers. The debate over whether central banks ought to be in that line of business has never been settled in favor of doing so.
Robert says yes. krasting says no. No resolution.
it seems a stretch to me to think that the Fed would ever really be concerned about benefiting the “real economy”…i am under the impression that their ZIRP & QE policies are still being directed towards recapitalizing the banks…you may have noticed that last week BofA announced they intended to institute a 5c a share dividend, and the Fed slapped them down, limiting their dividends to a penny…the same limit will apply to citigroup, but only if they do a 1 for 10 reverse stock split…those moves alone should tell us a lot more about the health of those two institutions than any passing grade on the stress test did…the banks are still playing Extend and Pretend , and continue to mark their assets to make believe…so the Fed’s monetary policies of zero interest and quantitative easing are to allow them to slowly recapitalize on the difference between their borrowing costs and the yields on their investments…last year John Hempton at Bronte Capital explained this in The arithmetic of bank solvency…
How about the Fed buying state and municipal bonds?
I have heard conflicting responses about the legality of that. But if it is legal, why not?
“Where would the Fed get the money to do all these nice things?”
The Congress has authorized the Fed to create money.
“I also don’t think that the Fed can drive up inflation expected for many years from now for the same reason. A higher declared inflation target should only affect inflation expected for the near future if the Fed will have some way of affecting inflation in the near future. It shouldn’t affect inflation expected for when the economy is back to normal — a promise to create inflation in the distant future by driving unemployment below the NAIRU is not credible.
So I don’t think that much more of the same will affect GDP much.”
This postings overall analysis reflects the continuing mind numbing Keynesian focus on the interest rate channel of the monetary transmission mechanism to the exclusion of all the others. But more importantly it is in serious denial with respect to QE2’s effect on inflation expectations.
The following numbers come from the data file of a Finance and Economics Discussion Series Working Paper called “The TIPS Yield Curve and Inflation Compensation” by Refet S. Gurkaynak, Brian Sack, and Jonathan H. Wright:
http://www.federalreserve.gov/econresdata/researchdata.htm
The estimated 2-year TIPS spread rose from 0.65% on 8/26/2010 (the day before Bernanke’s Jackson Hole speech) to 2.59% on 3/24/2011. Since the collapse of inflation expectations in late 2008 (it stood at -4.9% on 12/10/2008) the highest it was estimated to be prior to QE2 was 1.67% on 5/3/2010. This is the highest level that 2-year inflation expectations has been by this measure since 7/14/2008.
So according to Robert inflation expectations can’t be raised by QE2 and yet there it is, a 2% jump in 2 year inflation expectations since Jackson Hole.
Moreover there is evidence there may have been real effects already. According to the household survey employment is up by over 1.1 million, the most in any three month period (taking into account the year end population adjustment factor) in over 7 years. The unemployment rate just dropped by the most in any three month period in 27 years. And real final sales of domestic product rose by 7.1% in 20104Q, the most in 27 years.
And I am not alone in thinking this. Well known crackpot Martin Feldstein shares my opinion:
“To be sure, there is no proof that QE2 led to the stock-market rise, or that the stock-market rise caused the increase in consumer spending. But the timing of the stock-market rise, and the lack of any other reason for a sharp rise in consumer spending, makes that chain of events look very plausible.
The magnitude of the relationship between the stock-market rise and the jump in consumer spending also fits the data. Since share ownership (including mutual funds) of American households totals approximately $17 trillion, a 15% rise in share prices increased household wealth by about $2.5 trillion.”
http://www.project-syndicate.org/commentary/feldstein33/English
I guess the lesson in all this is that there are reality denialists on both ends of the political spectrum.
I don’t buy your assertion that the risk premium on Treasuries can’t go below zero. That’s only true if everyone’s investment return horizon is short. Institutions (insurance companies, pension funds, etc.) with longer-range liabilities have longer horizons. For them reinvestment risk is (at least on some margin that they will eventually reach) bigger than the risk of having to liquidate at a loss. The Fed can drive all the short-horizon players out of the Treasury market, and then the long-horizon players will be willing to hold longer Treasuries as insurance against reinvestment risk even if they expect a loss. Right now there is more than a 200 basis point spread between 3-month and 5-year Treasuries. And 5-year TIPS are yielding about negative 50 basis points. I think the Fed could, if it chose to do so, take that down by another 100 basis points and dramatically improve the incentive for investment in plant, equipment, durable goods, housing, etc..
Consider the thought experiment where the Fed buys up the entire national debt. For the sake of argument, let’s assume this doesn’t affect inflation expectations. Nonetheless, a bunch of entities that used to own interest-bearing securities will now own cash (on which the nominal interest rate is approximately zero). How plausible is it really that the bulk of them will still be satisfied holding cash? Some of them will, no doubt, but surely many will prefer to take a risk on assets that hold the potential for a positive nominal return. I can tell you, if my bond fund’s yield went down to near zero (without any commensurate weakening of the economic outlook), I would surely be moving more of my assets into the stock market, or into junk bonds, or into foreign assets. Even maintaining the assumption that it doesn’t affect inflation expectations, it’s clear to me that a POMO purchase of the entire national debt would have dramatic effects. A purchase of less than the entire national debt would have less dramatic effects, but on the margin (at least on some margin, when the Fed buys enough of the debt) it will have some non-negligible effect.
Having said that, I do agree with you that it would be preferable (offer more bang for the buck at less risk to taxpayers) for the Fed to buy risky assets directly. But there are some legal constraints, and I can understand the Fed’s political concern about not wanting to play favorites. If it were up to me, though, I would be going out as far as I legally could on the risk spectrum.
Two reasons not. One is, again, the risk involved in the municipal market. The very fact of a credit-easing goal, rather than quantitative easing (round 1 of Fed asset purchases, as compared to round 2) suggests the Fed is purchasing assets of questionable quality. Perhaps the Fed should be in the business of credit easing, but the tradition and theory (Bagehot-wise) has long been not to. The other reason is that “high-quality assets” generally means Treasuries, so there is never a question of the Fed directing credit. If the Fed gets into the business of purchasing assets with ideosyncratic risk – which is to say, pretty much everything but Treasuries – then it will be in the business of advantaging some economic entities over others. Again, that may not be a strong enough reason to leave munis out of the Fed’s portfolio, but it is among the main reasons the Fed does things the way it does.
I apologise for my stupid and rude reply to the comment — Sadowski’s paraphrase of what I wrote is fair and my recollection of what I wrote was incorrect. I withdraw my demand that Mark Sadowski apologise to me and I apologise to him.
I don’t find the argument convincing. In particular the numbers under discussion are generated automatically by a program related to a working paper published in 2008. The paper makes parametric assumptions about yield curves. I have no way of knowing if Refet S. Gurkaynak, Brian Sack, and Jonathan H. Wright still consider their assumptions reasonable. I will read the working paper and probably comment on it.
Giving money away is fiscal policy, not monetary policy. Indeed.
The Fed is only allowed to give money to wealthy banksters. We need to give money and jobs to those with unmet demand, not those who cannot find enough investments with acceptable risk premium.
kharris: “If the Fed gets into the business of purchasing assets with ideosyncratic risk – which is to say, pretty much everything but Treasuries – then it will be in the business of advantaging some economic entities over others.”
Thanks for your response.
But as for buying non-Treasuries, hasn’t that bridge already been crossed?
Your apology is accepted but I do wish to point out that I did not “paraphrase” you. That is a direct quote with all of the exact same letters, symbols, spaces etc.
Two years may or may not be “many years” but that makes little difference. Inflation expectations have risen sharply across the yield curve since late August.
And there are in fact other sources for 2 year inflation expectations:
http://www.bloomberg.com/apps/quote?ticker=USSWIT2:IND#chart
The two year inflation zero coupon swap rose from 0.92% on 8/26/2010 to 2.61% on 3/31/2011. Since the implosion of inflation expectations in late 2008 the highest it had closed prior to QE2 was 1.74% on 4/30/2010. In fact this is the highest level that 2-year inflation expecations has been since 8/4/2008.
So the story remains almost exactly the same using another measure of market based inflation expectations. I find little reason to distrust the estimates of inflation expectations derived from 2-year TIPS.