Friday, January 04, 2013
FOMC and Mr. Market
But the FOMC minutes were really interesting. I cannot think what the Fed was doing with that statement after that December meeting. I've never seen a more confusing set of minutes.
The FOMC statement in December was pretty simple - purchases of 85 billion a month, split between Treasuries and MBS, continuing until unemployment reached 6.5% or the inflation forecast exceeded 2.5%. Since the FOMC's released economic projections do not see either of those points being reached until 2015, it was reasonable for Mr. Market to assume that the Fed was promising to keep this up through most of 2014, at least. That was what the statement said, and it coincided well with earlier FOMC timing statements about monetary accommodation.
But bizarrely, the released minutes show that many participants thought that the program would have to be ended from mid-to-late 2013. It is not surprising that Mr. Market got a case of collywobbles after this news broke:
Participants also provided information on their views regarding the appropriate path of the Federal Reserve’s balance sheet. Most participants thought it was appropriate for the Committee to continue purchasing MBS and longer-term Treasury securities after completing the maturity extension program at the end of this year. In their projections, taking into account the likely benefits and costs of purchases as well as the expected evolution of the outlook, these participants were approximately evenly divided between those who judged that it would likely be appropriate for the Committee to complete its asset purchases sometime around the middle of 2013 and those who judged that it would likely be appropriate for the asset purchases to continue beyond that date. In contrast, several participants believed the Committee would best foster its dual objectives by ending its purchases of Treasury securities or all of its asset purchases at the end of this year when the maturity extension program was completed.So the majority, if I am reading this correctly, is in favor of ending asset purchases by the middle of 2013! How bizarre! How could anyone have understood this from the projections and the statement released in December?
Of course one problem, which is covered in the minutes, is that the Fed is getting in too deep with this policy. Since it is buying longer-term securities, when eventually it tries to unwind it is going to be in a rate trap. The value of the securities will have fallen and the Fed will take losses on selling if it needs to remove money to control inflation.
Central banks are banks, and they must remain solvent. It's foolish to assume that they will ever crash their own bank, no matter what they say. That's why ECB bond purchases are so problematic - should the ECB ever have to execute its announced plan, it would never be able to get out of it without some external source of funding to cover its losses.
Therefore I do not expect the Fed ever to sell the bulk of these securities. This implies that the Fed will instead have to use interest rate increases to control inflation, and raise rates higher than would otherwise be needed. Of course, when the Fed does that it closes off the avenue of selling the securities because it magnifies its own losses.
For banks and investors, this is a very confusing time. I think that stock traders must be quite dazed by being thrown this curve ball, but that the banking field had somewhat discounted the Fed's original statement. Bankers are looking at the strong possibility of much higher rates in a few years, and that possibility increases the longer the Fed continues adding to its balance sheet.
"Central banks are banks, and they must remain solvent."
Correct me if I'm wrong here, but the Fed is a bank that can award itself reserves out of thin air. It's the difference between "can't" and "shouldn't". All that is required is a "crisis" severe enough to dampen criticism of the move.
That, of course, would eventually destroy the currency. But 10 years ago I was told that the Fed couldn't bail out banks for potential losses taken on mortgages because THAT would eventually destroy the financial system. The last 20 years has been a process of defining banking deviancy down.
Good take on Fed losses. Really, all QE does is shorten the duration of the government's liabilities. The Fed+Treasury balance sheet should be presented as consolidated.
The portion of liabilities provided by the Fed in a consolidated balance sheet is traditionally of infinite duration (currency). With the IOR, now the lion's share has overnight duration (ER's). The more the Fed substitutes its liabilities for term Treasuries, the more overnight duration ER's are issued to fund the government.
Bottom line: the more QE, the more tax payer is exposed to rate increases; of the more the Fed is forced to issue more liabilities to cover losses, thus losing control over base money. Either way, its a function of duration risk.
Either path forward implies a loss of future power of action by the Fed, which implies both real rate increases and the consequent loss of a degree of control by the Fed.
When you look at government financing, especially consolidated with Fed ops, it's clear that the US has taken on massive rate risk for its obligations.
Hardly reassuring, but that is why I don't think the Fed can ever sell most of these securities.
I look at this situation, and I think "unstable system", and so I think the Fed's priorities will shift toward trying to maintain some degree of stability as the first priority within two to three years.
Of course, this has everything to do with bank lending policies right now and in the next couple of years. You can't have money tied up in long-term fixed-rate obligations.
It also implies the pending taxpayer liabilities with FHA and so forth are higher than we think.
But even if they don't sell the securities, doesn't the central bank (ECB or Fed) have to mark them down to market on their balance sheet? And doesn't this automatically create insolvency?
That's not to say that an ordinary bank doesn't have to be very careful about getting caught in longer term securities when their short-term cost of funds goes up past the investment return on those securities.
That's the reason that the ECB is so adamantly saying "no" to the Greek bond writedown for the public holders.
OK, so I'm confused. When is a central bank like a bank (concerned about taking a loss on assets it holds), and when is it like a sovereign (using essentially the seigniorage power to create reserves out of thin air)?
This is a good question.
Forget about seigniorage. The Fed pays out interest on reserves. Right now it makes a spread b/t that IOR and interest income on Treasuries/MBS. If the IOR rises to contain inflation, that spread becomes a loss.
At that point, the Fed can do three things:
1) lower the IOR -- even to zero -- to eliminate the losses. This would cause inflation to accelerate.
2) get the taxpayer to cover the losses. Not fun politically.
3) credit banks with the interest "out of thin air"; that is, by issuing them reserves without asking for Treasuries in return.
So, basically, options 1 and 3 are highly inflationary. Option 2 puts the taxpayer on the hook.
Economists always assume that options 1 and 3 will not be necessary because all the Fed has to do is briefly raise the IOR and inflation magically disappears. There is no room in their mind for the Fed getting "behind the curve", where it raises the IOR but not enough to stem inflation, then does it again, then again, etc.
The purpose of most modern central banks is not to make money, but to keep at a minimum, price stability, and in many cases to maintain their own currency within certain ranges against other important world currencies.
Withdrawal of money can be done through various mechanisms. For example, the central bank could sell securities it holds, which then withdraws free money from the system. Or the CB could raise its own deposit rate, so that banks could make more money by leaving money on deposit at the CB, which would tend to withdraw money from the marketplace. Instead of lending it or buying securities, banks would instead leave it on deposit at the Fed.
The problem is that if a CB gets its balance sheet very loaded, past choices begin to constrain future choices, sometimes to the detriment of the future.
It may be theoretically possible for a CB to manufacture a bunch of money to cover its losses and continue operations, but in general manufacturing money cannot constrain inflation, so if that is the future need, the ability to print does not get you there.
In order to have scope for these activities, the CB needs to have a certain level of assets on its balance sheet. For example, most modern CBs are going to buy at least 10% of their country's sovereign bonds. But if the Fed were really to buy 85 billion extra worth of bonds each month while continuing to roll the proceeds from maturing securities for a year and a half (18 months), that would add 1.5 trillion worth of securities to the Fed's current 2.67 trillion, of which less than 90 billion are inflation-hedged. That is a lot of rate risk or duration risk!
Shifting that risk to the taxpayer, if we were paying our bills without borrowing, would tend to withdraw money from the system. But we are not, so it would probably just raise the deficit, causing more borrowing.
Hmm, this really deserves a long series of posts, doesn't it?
Other methods used to control inflation are increasing required reserves for banks, which pulls money out of the lending pool and contracts credit and the money supply. India's CB uses that tool very heavily.
It is clear that under current conditions, with NIM (net interest margin) so low, the Fed doesn't have that much scope to use this tool. There are vulnerable institutions. Also, many banks have excess reserves already (or claim they do), so the implication is that the Fed would have to raise reserve requirements greatly to accomplish much, which expands the list of vulnerable banking institutions.
Neil, you might be interested in this post by an MMTer. By which you can see that I believe that some of the MMTer critiques of more mainstream theory are quite correct, and in any case it is a good discussion of the subject that is less US-centered than most.
The actual relatively wide variance in CB operations even in modern times is a very interesting topic of its own.
I'm going to have to think about your last comment for a while.
The thought I have now is that the Fed's monetary actions are being constrained to an ever-greater extent by fiscal policy. After all, if the federal deficit were smaller then the amount of QE required to keep rates down would be smaller as well, am I right?
In the last 20 years we've already seen a few occasions when the Fed abandons its price-stability and unemployment goals in order to protect its unstated (but primary) third goal of maintaining the financial system essentially as JP Morgan left it to us. The hitherto-unprecedented policy tools put in place in each of those crises seem to quickly become SOP. The Fed is getting good at thinking the unthinkable. What I've been looking at is largely what CAN be done when the unthinkable becomes the only choice.
It would probably be instructive to war-game this from a central bank institutional perspective, though.
Links to this post: