“Not yet, Bank of America,” says the Fed, “but the rest of you guys go right ahead. Yes, even you, Citi.”
Thus the second round of stress tests came to an end with the conclusion that key players in the banking system are sufficiently well-capitalized that they can be allowed once again to pay dividends to shareholders.
The economists’ debate about bank dividends has so far been between those who see payment of dividends as a drag on recapitalization through retained earnings (as Admati), and those who see the same payment of dividends as a first step toward acceleration of recapitalization through stock offerings (as Isaac).
Now comes the Fed itself, announcing its own record dividend payment to the U.S. Treasury. $79.3 billion is a lot of money, even on the Fed’s swollen $2.4 trillion balance sheet. We are talking 3.3% of total assets—which is apparently what you get, after expenses, when you borrow at near zero percent and lend at near 4%.
This same carry trade is of course the origin of the recapitalization of the private banking sector (albeit with somewhat higher “expenses”). The difference is that the dividend is paid to private shareholders, mostly in the form of increased stock valuation, but now also in the form of actual cash dividends.
Looking forward, maybe the banks will be required to raise more capital, so diluting existing shareholders. But the more significant dilution has already happened, namely dilution of the taxpayer interest. Indeed that was the whole point of the exercise—to restore the buffer of private capital that stands between bank losses and the taxpayer.
From a money view perspective, the interesting issue raised by the entire exercise is not so much whether this is the best way to raise bank capital, but rather the prior question, What exactly do we mean by the concept of bank capital?
What we mean by bank capital is the difference between assets and liabilities, which is to say the difference between the present value of future cash inflows and the present value of future cash outflows. But what value should we put on these future flows? It is a familiar problem in corporate finance, but banking presents a special challenge because of the greater relevance of the survival constraint.
Moment by moment, the significant constraint on a bank is not solvency but liquidity, meaning the ability to meet promised cash outflows with actual cash inflows. This “survival constraint” binds today, and also tomorrow and the day after that, on into the future. This has consequences for valuation.
Consider the valuation of a bank with and without a guaranteed source of refinance. With the refinance guarantee, we know the survival constraint will never bind at any point in the future, so we can ignore it for valuation purposes. But without the guarantee we have to make a judgment about the probability of hitting the survival constraint, not just tomorrow but on into the future; present valuation will be lower.
It is of course entirely possible that, for any given banking entity, valuation is positive or negative under both assumptions. Those are the easy cases. Sometimes I think the regulatory reform process is driven by the dream that somehow we can fashion a world in which these easy cases are the only cases!
The hard case is the one where bank valuation is positive with the refinance guarantee, but negative without. Refinance guarantee, and nothing more, makes the difference between solvency and insolvency.
From a money view perspective, the whole problem with the debate about bank capital is that it does not adequately appreciate that this hard case is not a special case, but rather the general case. The whole business of banking inherently involves making a specialty of dancing close to the edge of the survival constraint, so that the rest of us can choose a different specialty. For banks, refinance always matters because sometimes it is the only thing that matters.
In boom times, it doesn’t seem to matter much, since everyone has access to the world money market. Asset values thus come to reflect the idea that the survival constraint does not bind, that refinance is guaranteed, that liquidity is a free good. But then comes the crisis, and suddenly it does matter. Access to the central bank is everything. If you have access to the central bank, then you are liquid, then you are solvent; if you don’t, then you aren’t, then you aren’t.
The fundamental issue, from this point of view, is not capitalization but rather access to central bank refinance.
To be continued…
