The blogger Felix Salmon, at Reuters, was kind enough to link my post from this morning about the Bair-Miller-Moore Haircut and to disagree with it. I will try to answer Mr. Salmon point-by-point and also offer a more general, conceptual response.
He writes that the haircut, or the threat of one, "means that lenders to dodgy banks will actually have to start doing underwriting, rather than simply relying on their security interest. That keeps everybody honest, and will give the system a heads-up when banks start getting into trouble." In fact anyone lending to anything or anyone "dodgy" – a bank, a business, a homeowner -- will either want collateral or will want to be paid higher yields. If collateral is unavailable, the yield will go up. That is why the interest on your credit card (no collateral) is 20% or 30%, while the interest on your home mortgage or car loan is 5% or 7%. The Bair-Miller-Moore haircut means less collateral for those lending to banks, which means the lenders will look for increased yields to compensate for the lack of collateral. The banks are going to pass those costs along in the form of higher rates on loans, which is counter to the goal of economic recovery and getting credit flowing again.
He also writes that the haircut "means that wholesale lenders no longer have the ability to jump the queue when it comes to seniority. Banks should repay their depositors first, and then their senior unsecured creditors, and then their subordinated creditors, and then their preferred shareholders; whatever's left over goes to common shareholders. But increasingly the pecking order has been upended by allowing banks to issue secured debt, which in practice ends up being senior even to depositors. In some countries, banks aren't allowed to issue secured debt at all." What countries are those? It's a good bet that in those countries, consumers have to pay higher interest rates on loans and earn lower interest rates on deposits than they do here in America. Kind of like the situation in Canada, where other factors also come into play. And if Mr. Salmon is going to mention depositors, how about this for a thought experiment: Imagine the panic that would ensue were Sheila Bair to announce that henceforth, depositors at FDIC-insured institutions would lose $20,000 of each $100,000 in insured accounts if their bank went under. That would have the virtue of, as Mr. Salmon puts it, giving the system "a heads-up when banks start getting into trouble." It would be a heads-up in the form of a bank run. There may be a theoretical case for such a policy, but as a practical matter, it would undermine confidence and reduce the deposit base at a time when regulators are trying to restore confidence and increase bank capital.
What's more, as a practical matter, the "problem" that Bair-Miller-Moore seems aimed at solving is a nonexistent one. The losses at banks that have flowed through to the FDIC are not due to repo lines – the secured debt of the sort that Bair-Miller-Moore would impose a haircut on. Where the FDIC has lost a lot of money – at Indy Mac, for example -- there were not any secured creditors outside of the FHLB. Most of the repo market is going to high quality banks and the broker dealers. Lehman had a lot of repo. But how much did the FDIC or other government branch lose on Lehman? Zero.
Mr. Salmon suggests that banks raise capital through equity instead of through debt. Using his example of a $1 million asset, if a bank is going to issue equity stakes for $1 million to fund a $1 million loan, the provider of the equity is going to say, "hmm, if I give this bank $1 million of equity, how much should I make?" The prospective investor can buy Con Edison common shares and make a 6% dividend plus capital appreciation, so let's say, 10% on ED stock. The prospective equity investor will want to get at least 10% for equity in the bank. So maybe he'll say, okay, I will give the bank $1 million, but I want a 10% return." To generate that return for the equity investor, the bank will need to earn in excess of 10%. Banks that finance themselves this way will have to charge mortgage rates of at least 10%. Good luck with that. Mortgage rates are currently 5%. So if I borrow $200,000 at 5%, I can afford to pay $10k of interest per year, or $833 per month. If mortgage rates go to 10%, I can still only afford $833 per month, which means that I can only borrow $100k, which means that if previously there was demand for homes at a home price of $240k, now there will be demand at $140k. So, home prices will tank. Is that the policy objective of Mr. Salmon, or of Bair-Miller-Moore?
Mr. Salmon's response to the danger that repo funding for banks will dry up is a somewhat flippant, "they have access to the Fed's discount window anyway." Great. The repo markets are basically private, market-driven contracts. Do we really want to replace their role with that of the Federal Reserve, a government agency that can barely achieve its statutory assignments of maintaining the value of the dollar and a reasonable unemployment rate, let alone financing the banks? And here we thought the idea of Bair-Miller-Moore was that it was supposed to make federal aid for the banks less likely, not more likely. If a bank's only source of liquidity is the Federal Reserve, then we become dependent on the Fed to determine the price of assets. Having a healthy repo market means that third party private markets determine what is safe in terms of assets, and where they should be priced – far better than depending on the Fed to do the job.
Mr. Salmon claims that the rights of secured creditors did not take a beating in the Chrysler bankruptcy, and that the creditors merely "sensibly didn't exercise those rights." But Richard Epstein, who is an authority on property rights and the rule of law, says otherwise. There were all kinds of tales of threats at the time, including an on-the-record claim by the lawyer for the creditors. When one of the Chrysler bondholders caved, they issued a statement attributing the judgment to "considering the President's words." There were people running around the White House with the title auto "tsar." When the president of the United States is going around telling you and the rest of the country that if you exercise your rights you are a greedy speculator putting your own interests in front of the nation's, it isn't much of a right. What Mr. Salmon attributes to "sense" is more accurately attributed to the blunt threat and use of government power and intimidation. Or, as Amity Shlaes put it wonderfully last night at the Manhattan Institute's Hayek Lecture: "The administration made a political decision to steal. Steal for labor from the owners of the bonds."
Mr. Salmon asks: "What are banks doing holding so many securities in the first place? Shouldn't their assets mainly be loans, which can't be reposed?" In fact the major security banks own are residential mortagage backed securities wrapped by Fannie and Freddie. Those securities help make mortgages available to average Americans. If you want to eliminate them, fine, but good luck getting a 5% mortgage from your local bank if they have to hold it. Look at the difference between jumbos and conforming loans – conforming loans cost 4.9% and jumbos cost 6%.
The broader issue is that Ms. Bair, and Mr. Salmon, and FutureOfCapitalism.com all tend to agree that having taxpayers bail out banks is a bad idea. One approach to the problem -- the Bair-Miller-Moore approach -- is to try to tinker with the rules about the private funding sources of the banks so that the banks are less likely to require government bailouts. That has the disadvantage of being likely to impose higher costs on consumers and businesses. It also might not succeed in preventing bank failure. And it also restricts the banks' ability to contract with other private parties, and of other private parties to contract with them, which is a freedom that is important in capitalism.
My alternative preferred approach to the problem is to try to leave the banks the freedom to contract with creditors however they choose to, but just not to have the taxpayers bail them out if they fail. It's the implicit government guarantee on the too-big-or-interconnected-to-fail banks that is the source of this problem in the first place. If there's an understanding that the government is going to rescue a big failing bank, then it's easy to follow from that logic that the government can try to restrict what they do. Without that implicit guarantee, it's much, much harder to justify the restrictions. Update: To clarify, the guarantee I am talking about is TARP-like capital injections and subsidies through the Temporary Liquidity Guaranty Program. The FDIC guarantee to depositors is another matter; for my thoughts on that, see here and here.
