Quantitative Easing without the Fed
Robert Waldmann
Alternative title: Stimulus without Congress.
Many argue that the Fed should buy long maturity and/or risky assets in order to reduce interest rate differentials. The Federal Funds rate is essentially zero, but interest rates which matter are still significantly positive so more could be done by some huge player in the bond market.
But why the Fed ? The Treasury is a huge player in the bond market. They are still selling long term bonds. Why ? What if the Treasury decided to finance the deficit with 1 and 3 month T-bills alone ? A now deceased parrot said that this would reduce the slope of the yield curve.
But so long as the Fed keeps the target federal funds rate roughly zero, that means lower medium and long term interest rates.
There must be something wrong with this proposal. What ?
What is the profound difference between the Fed buying more and the Treasury selling less which outweighs the fact that Obama can fire Geithner and can’t fire Bernanke ?
The obvious difference is the percent of tax revenue that goes into interest and debt payments goes up with your proposal. Consider the $1T of agency mbs purchased by the Fed: lowers Treasury risk, pumps money into the market, and then contribute $70B/year to Treasury. What should happen is that the government figures out how to sell DOE loans to the Fed, allowing DOE to increase its portfolio of energy investments without allowing the Republicans to stop it. That would be stimulative
Good question. Here’s a partial answer:
1. For any entity that cannot print money, and which lacks a lender of last resort, it is dangerous to borrow short and invest long, hoping to roll over the debt as it comes due. Witness bank runs, not to mention Greece and Spain. So governments don’t like to finance their lon-term borrowing requirements with short term bonds. Now it’s true that the Fed would (almost certainly) act as LOLR to the US government, but it would be a lot easier if the Fed just did the job in the first place.
2. QE itself is of little consequence, if it’s assumed to be temporary. QE works mostly by signalling the Fed’s intention to raise inflation above where it’s currently expected to be. And if inflation is higher, then the QE must be at least partly permanent. Responsibility for inflation is assigned to the Fed, not Treasury, because ultimately the Fed has control of the money supply.
Damn, but the print in this comment box is small! I take no responsibility for typos, since i can barely read what I’m writing!
What’s the intent of this round of QE? Is it to drive longer term rates lower? If so, could this have the opposite effect of freeing up private capital for investment?
“There must be something wrong with this proposal. What ?”
We did pro-cyclical policy already. Been there , done that.
Average duration on Treasuries is 4 years, due to short range financing the past 10 years. Largely because their central bank buddies like 3 year or less maturity. Timmay recently said we need to go longer, because $9T and up is getting scary from the standpoint of a potential interest rate spike on refi needs. At the moment the 5 year T-note yields 1.25%, so that is almost same as cash anyway.
This also sheds some light on the wisdom of the “increase inflation target” solution that Very Serious Economist’s ™ have proposed to lead us out of the Liquidity Trap and near deflation numbers. Guess what happens to government finance? The japs didn’t go for that proposal either when it was made to their policy makers.
However, they may need to buy all the MBS now because we are finding out they may be no good. Otherwise, look for spreads to widen here.
And if our un-nationalized banks are supposed to earn themselves back to solvency, flat lining the yield curve and spreads seems an odd way to do it.
I think we’ve played out our two big macro tools for all they are worth.
My proposal was same total debt issue but different maturities. Interest payments would decline.
I guessed that there was some prudential rule about how the US Treasury should match maturities. But I think that’s totally silly. The US Treasury is the US Treasury not the Greek Treasury.
I don’t see how the Fed can signal it’s willingness to have higher inflation when we are no longer in a liquidity trap, with or without QE.
Thanks for the thoughtful comment.
Hmm the banks seem solvent. At least investors seem to think they are. Procyclical it would be once we are above the trend line. For now, we have a huge output gap and output far below potential.
Refi no problem if the Fed keeps the short term target interest rate low. That implies that the Fed would refi the Treasury. That was my caveat.
I’d say the issue is that Geithner (and all secretaries of the Treasury before him) think the Treasury should act like a firm. I don’t think this makes any sense. We want huge deficits when there is trouble. That’s an automatic stabilizer.
I don’t see why the balance sheet of the Fed vs that of the Treasury matters so much. Note if short term interest rates spike, the Fed would lose a lot if it buys long term bonds as part of QE. That isn’t so different as the Treasury bearing the loss as it refinances.
I think the idea that sound policy is to pretend that the Treasury is a big bank is always foolish and is foolish in this case as well.
Larry Kudlow thinks banks are solvent, but others think relaxing mark to market rules has a lot to do with it. And even if some are ok, it’s still a mixed playing field. And getting worse again with the MBS servicing situation.
Re: Procyclical: I was making an attempt at econ humor. I meant during the boom years we had both procyclical fiscal AND monetary policy, ie shortening up maturities that the Treasury SOLD.
“Refi no problem if the Fed keeps the short term target interest rate low. That implies that the Fed would refi the Treasury. That was my caveat. “
So…the FOMC does 9 to 20 trillion in revolving T-Bill purchases with the Treasury? No problem? Ok, I guess I can’t say for sure. Maybe that will be good for employment if they don’t automate it too much.
The Fed gives up its ability to do monetary policy when it has too much long term stuff on the balance sheet, because they aren’t able to retrieve all the liquidity they put out there whenever they decide they need to. Whether it matters is all about confidence in the system. The US as the reserve currency is supposed to set the standard, and it is becoming a pretty poor standard.
If we get the interest rate spike or just a slow increase, interest costs mushroom for the treasury, and we have real stuff we would like the USG to pay for.
The problem I see is we have been abusing the auto stabilizers, using them to create unsustainable booms and tax cuts, rather than pumping them full of air again for the next downturn. Now all we have is a lowrider.
Also, the reason the Treasury can borrow money was because private investors and foreign CBs think that we, the taxpayers, are the the assets of the Treasury Bank. We don’t really have any experience in this country how any new way would work.
Robert –
When did we leave the liquidity trap?
Nick –
On many browsers, {Ctrl}{+} will magnify test in the active window.
Cheers!
JzB
Well, if I knew jack about econ, I could juxtapose this with what Karl Smith posted today, and come up with a plan.
http://modeledbehavior.com/2010/10/10/thats-so-money/
Oh, hell – why let a little thing like that stop me. Suppose the the Treasury were to buy huge wads of commercial paper to drive down the interest rate, which suddenly diverged from the FF rate in fall of ’08. (Has that EVER happened before?)
Coincidently, this happened at exactly (near as I can tell) the same time that the M1 multiplier went from a long, slow collapsed into catatonia. No signs of life any rime since, either.
What are we to make of that?
Yours in perpetual confusion,
JzB
s/b:
M1 multiplier went from a long, slow decline into collapse and catatonia. Still no signs of life.
JzB (bad cold, poor proof reading skills)
The liquidity trap is the essentially zero level of safe short term interest rates. Long term interest rates are not essentially zero. The 30 year rate is around 3%. The (ok maybe an) idea of quantitative easing is that the Fed can drive down long term interest rates by buying huge amounts of long term bonds in exchange for money or 1 and 3 month t bills (which are trading as perfect substitutes).
Term premia, which sure aren’t zero, might be less important than quality premia. To drive down returns on risky assets, the Fed (or the Treasury) would have to buy risky assets. The Treasury clearly has the legal authority to stop selling long term bonds (or in Treasury speak “bonds”). It does not have the legal authority to buy risky assets. So my plan would not touch quality premia.
My view on others’ view on banks is based on asset prices. Their shareholders sure think they are solvent. I do not think that financial markets are efficient so I don’t claim that banks are solvent. But their equity is priced rather far from zero.
no cold and didn’t notice any problem with the original sentence. I don’t think that FF and commercial paper ever diverged by much before. The Fed can buy huge wads of commercial paper, but I don’t think that the Treasury could without authorization from congress. The point of my proposal is that the Treasury can and does decide what debt instruments to sell on its own.
The buy huge wads is the quantitative easing that many people are begging the Fed to do. IIRC commercial paper yields are pretty normal. In fact, I think that there aren’t any really high yields this side of toxic sludge (low rated corporate bonds have a lower yield than in 2006). However, even lower yields on corporate debt might encourage investment.
Note “might” it is also possible that given excess capacity that no real interest rate above minus 5% or something will cause much of an increase in investment in which case QE wouldn’t work.
Robert,
Have you changed your general opinion on quantitative easing since the last time (Nov 09) you posted on this subject?
Treasury has committed itself to extending the average life of the debt. To do that it needs to sell more long term bonds. Not less as suggested here. If we funded an increasing portion of our debt with T bills we would become much more subject to a disruption in the market. Too much ST debt = death.
Gentlemen & ?, thank you for the convoluted discusion this morning. I feel much better now. Gee, the world as we know it can continue on its merry way.
Basel III has a 7-10 year phase in for new capital requirements, which aren’t that tough, so that’s enough of a hint for me.
My definition of bank solvency wouldn’t be if the FDIC decides to shut them down. I’d say it’s capital adequacy that could stand some shock. Like maybe the Fed first flatlining yields, then if we do ever get inflation/interest rates to rise, mark to market balance sheets contract. Ouch.
shorter term interest would be higher, no?
I do have a question that maybe you can answer: what was the rationale for the US ban on bank control of non-bank enterprises and when was it put in place? It’s interesting to see how the major components of US manufacturing were all financed and even operated by core banking enterprises at one time. Without mellon bank, the midwest manufacturing sector would have been very different.
I’ve been struggling fruitlessly to understand what it is that is puzzling you.
It seems like your model has to be one of supply and demand for loanable funds, controlled by a market interest rate. The Treasury sells a boatload of long-term bonds, and it drives up interest rates, as markets clear only at a higher interest rate; or, the Treasury sells no long-term bonds and finances its need for funds with a boatload of short-term bills, driving up short-term rates, except the Fed doesn’t allow short-term rates to rise, and the dearth of Treasury “demand” for loanable funds, aka long-term bond issues, causes the long-term market rates to decline.
And, you are wondering why the Treasury and Fed acting together to finance a very young national debt, at Treasury’s initiative, is not considered equivalent to the Treasury managing the age of the marketable national debt according to prudential rules of some kind that advise trading away savings on debt service in the short-term, for a reduction in interest rate risks and turnover challenges, while the Fed, say, buys boatloads of long-term Treasury bonds, driving down long-term rates, as part of a policy of “quantitative easing”, whatever that may mean.
Your alternative title, “Stimulus without Congress”, trips me up. Are you trying to puzzle out “quantitative easing” and its channels of economic effect? (Aren’t we all?)
I find myself confused by your confusion.
If you really are using a model of loanable funds flowing through a market with a market-clearing interest rate, and that’s not just my misunderstanding what you’ve written, then maybe that’s the problem. The Fed isn’t managing the flow of loanable funds, but, rather, the unit of account and medium of exchange, underlying the flow of funds. Also, in important ways, there is no hydraulic flow of loanable funds to begin with; it’s mostly intermediation and carry and credit. In this scheme, the marketable national debt is in the nature of a public good, providing a risk-free reference for calculation and a means of insurance for hedging. For that purpose, the marketable national debt should be spread out over the whole range of terms: a steady torrent of taxes, borrowings and repayments that helps to stabilize expectations about the demand for money.
The Treasury is constrained to borrow what the Federal government needs to finance its operations and purchases. Formally, that constraint relates to Congressional appropriation authority and debt ceiling authorizations, but it seems to me, intuitively, that constraint is inherent in the role and function of the Treasury, not the legal particulars. The Treasury might manage the term structure of the marketable national debt, but it’s not clear to me that they have much discretion about how much debt there is to be. (The wonk in me wonders about Treasury discretion over the non-marketable national debt, which is very large; there might be some strategic discretion hidden there.)
The Fed, as a central bank charged with managing the marketed national debt, and with it, the money supply, inflation and the level of economic activity, is not constrained by the particulars of the Federal government’s funding needs or Congressional appropriations.
The problem with having the Treasury sell only short-term bills, as a way of starving the long-term bond market into lower rates can be viewed in two ways. One is that it imposes the constraint that the Central Bank accomodate the demand for short-term funds to keep short-term rates from rising. The constraint removes the discretion that gives the Central Bank control of the money supply.
Another way is to see that there can be no […]
I think we need to give at least some credit, and thought, to what institutions say they are doing, and why. Treasury does claim to be engaged in debt management, rather than in economic management, in deciding the maturity profile of its debt. Roll-over risk is one issue in debt management, and presumably a small one, as you suggest. There are other issues, though. Funding at short maturities involves interest rate risk, as well as roll-over risk. Treasury seems pretty dedicated to serving its benchmark function.Treasury surveys its client base to see what the client base wants, and they have apparently told Treasury they want long-dated paper. That is partly a market-efficiency issue.
If we consider just one goal, then yes, Treasury could do what the Fed is doing as regards yields – maybe. Treasury seems to have more than one goal, and a Fed that has the institutional responsibility for interest rate policy.
I’m convinced that there is some announcement effect and supply effect in the short term regarding the shape of the curve, but what does theory say the curve should do if long-term liabilities are financed with short paper? Isn’t there some presumption that long rates should reflect liabilities, as well as supply? A term structure sort of notion?
Your points are well taken, Bruce. Unfortunately, it appears that Robert blew off the comment thread.
Would you have any interest in writing main posts for Angry Bear? You’re a great thinker and writer. Nice to see you participating here.
The only relevant difference is in the balance sheet of non government.
Treasury term structure change offers a different choice for treasury term structure.
QE offers no such choice in that it creates M1 bank balances for end sellers of treasuries.
The question is whether that is significant. Not for short term interest rates, but perhaps slightly for risk. It depends whether institutional money managers will use the same thought process to determine what to do with those balances as they do when choosing between short and long treasuries – i.e. will they buy short bills or something else when they sell their long treasuries?