Fed to bond bears: we made $98.7 billion in 2014. How was your year? http://t.co/8xV506VDNf
— Brian Romanchuk (@RomanchukBrian) January 10, 2015
(I find this outcome as being somewhat entertaining, as my "tweet" above suggests. I was at a conference in the early days of the programme, and I got stuck in a conversation with an earnest analyst who explained how worried the Fed was about getting stuck with losses on their portfolio. I told him that the Fed was arb-ing the market and would make out like bandits, but he refused to believe that I knew what I was talking about.)
These Fed profits represent roughly 0.6% of GDP. If the Fed had kept its portfolio at a reasonable size, invested in Treasury Bills, and paid interest on reserves, the profits would have been pretty close to zero. Fixed income trading profits are close to a zero sum game; this means that this policy has lowered "private sector" (if you want to include foreign central banks as part of the "private sector") income by 0.6%.
To what extent QE lowered interest rates, "private sector" interest income from the Treasury was lowered further. (There would have been a reduction of interest costs for other borrowers, but that nets out with the reduction of interest income that would have been received.) Since I do not think QE lowered interest rates by much, I do not think this effect was that large.
The supposed benefit for the economy is that lower interest rates are supposed to stimulate activity. For example, it may encourage businesses to borrow, or probably more importantly, encourage households to borrow in the mortgage market. Although I see that this effect can help, it is somewhat conditional on the state of the economy. With the U.S. housing bubble now burst, I see much less of an economic boost from lower interest rates. But even so, it is hard to see the benefits from this "channel" being much larger than 0.6% of GDP, given that the reduction in long-term rates was probably less than 100 basis points.
Profitability Does Not Matter, But...
It makes little sense to worry about "profitability" as a concept when discussing the Federal Government; it creates money at almost no real cost (other than printing and minting costs, but most government money take the form of electronic book entries). What matters for the government are real resources.
However, it is necessary for the Fed to control its expenditures for reasons of political accountability. It cannot have employees running around buying things at random on the theory that "money is free for the Federal Reserve". Correspondingly, costs have to be controlled, and it is reasonable to expect the system to run with a slight profit (assuming that it does not extend duration, which is what it has done during this QE episode).
See Also:
(c) Brian Romanchuk 2015
I wonder whether extensive QE is useful to: 1. put a price floor under certain financial assets; or 2. help banks recapitalize; or 3. provide banks and nonbanks w/ability to swap floating NAV assets for fixed NAV assets (swap investments for money) which relates to establishing the price floor in 1. If the tax on income is offset by a gain in financial asset price stability then QE pays for itself with the hidden benefits.
ReplyDeleteFinally, I ask, where are all the bad loans being written-off? Where is the so-called "bad bank?"
So maybe QE helps banks avoid taking a large loss all at once and lets them write-off bad debt over longer periods of time against profits sponsored by Fed/Treasury cash flow assurances. This is part of the recapitalization theory per item 2 above.
The first wave of QE was the purchase of risky assets, which is the same thing as a lender-of-last-resort operation. It certainly helped unfreeze the markets. The later rounds of QE were the purchase of GSE-guaranteed mortgages and Treasurys, and just increased the monetary base. This did little for the private sector.
DeleteThe bad loans were largely securitised. Most of them should have been written off by now, and the bondholders will have eaten the losses.
Fed provided borrowed reserves to troubled banks in late 2008 as lender of last resort and at that time sold Treasuries to neutralize reserve levels and control interbank fed funds interest rate. This policy would have caused a problem if the trend were sustained. Fed would have run out of Treasuries to sell. In 2009 Fed purchased MBS which I think were all products insured by GSEs and backed by the Treasury bailout. So Fed took on interest rate risk in MBS but not so much credit risk. By purchasing securities from banks Fed adds reserves, and if nonbanks sell to Fed and banks clear payment, then Fed adds reserves and deposits when it purchases securities from a nonbank. So QE is similar to and different from lender of last resort via expansion of discount window borrowing exclusively by banks.
ReplyDeleteBanks had sold securitized loans to nonbanks with credit enhancements. So the write-off would not just occur in nonbanks, some of the bad loans were bouncing back onto bank balance sheets. This expected write-off would be a charge against bank capital and caused problems for banks in money and capital markets at the time when they needed to develop more liabilities and capital. Fed/Treasury policy with QE solves some of these problems for the aggregate bank but does not involve the government in direct allocation of the write-offs as far as I can see. Either the govt or nonbanks must put up capital to charge off a bad loan on the aggregate bank, but nonbanks tend to run the banks when a large capital charge off occurs, so government provides the capital, or the loss is strung out across future revenue via regulatory forbearance.