Viral Acharya, David Backus, and Raghu Sundaram (teachers at NYU's Stern School of Business, whose work is part of the NYU Stern project, “Repairing the US Financial Architecture: An Independent View”), writing in the FT on January 6th, employ a popular, but naive, logic that banks ’short of capital’ or ‘undercapitalised’ are 'insolvent' more than 'illiquid', as shown by the fact that central bank bail-outs are what let the banks pay generous dividends earlier when they should have instead been topping up reserve capital in anticipation of losses ahead. This argument is not new. 
The counter-argument is that banks are not like ships with detailed charts of the dangerous reefs and choppy water ahead. They could not collectively or individually know the timing or systemic scale of credit risk and market risk losses yet to come. One may argue they should have had a pretty good idea - but that is another argument entirely. Anyway, whether they should or not, they clearly did not. Also, not any banbk can get government help. Only systemically important banks are helped to recapitalise (restore their capital reserves) by central banks. And when banks are systemically important it is not just to each other, but to the whole economy.To be short of capital does not mean insolvency in terms of zero or negative net ‘book value’ (liabilities exceeding assets or vice versa). Much criticism has been levelled at banks having loans (assets) too far in excess of deposits (liabilities).
But, that is not at all a simple measure and may not be any kind of weakness beyond indicating (but not confirming) a possible funding problem. Neither does it mean the bank was previously under-capitalised. All that being short of capital means is that loan-loss provisions triggered by loan defaults (before recoveries) were ‘unexpected’ and (on paper) have reduced capital reserves below the regulatory minimum at a time when the cost of interbank borrowing is prohibitive. In my estimation, based on Bank of England and Federal Reserve data, a normal recession wipes out most bank reserve capital. This time (with the credit crunch additional effect) banks' reserve capital is being wiped out twice! Government help is replenishing one times bank reserve capital. The banks still have to find ways of replenishing as much again. We have to allow for times when replenishing capital reserves is not easy. The steep rise in interbank borrowing costs was unexpected. This is not just a market price, but also an OTC bilateral price. Banks were wary of the reputational risk of signalling ‘dangers ahead’ by cutting dividends and thereby triggering further loss via ratings downgrades. This was especially so for banks on the cusp of losing unsecured borrowing credit status.
The argument by Acharya, Backus, and Sundaram may be reasonable if only very few banks are in difficulty and have very low, zero or negative ‘book value’. To say that “There seems to be little in place to stop an insolvent bank from raising cash against shoddy collateral” is clearly wrong or they would all be doing it. To say, “Central banks should refuse loans to banks, and governments to other firms, that cannot demonstrate their ongoing viability” is also clearly wrong if their viability is in the hands of the central bank and the government. Furthermore, the viability of government is itself at risk if it fails to support the banks and major industries like US Auto. This is not a two-dimensional zero-sum set of problems. To say “Government money should not be a gift to existing investors and management” is wrong as there are no outright ‘gifts’, and government funding is not cheap! To say, “Without such strings attached, government help is likely to be more expensive and, perversely, reduce the willingness of the private sector to contribute to its own survival” is obviously wrong to since there are many penalties involved in publicly receiving government aasistance, only a few of which are explicit - the rest being dictated as severe punishment by the markets.The classic literature worries about ‘runs on the bank’ by depositors. But, shouldn’t we worry as much about banks mounting ‘a run’ on all their borrowers? If we worry about whether dividends should’ve been paid before bank shares fell 70%, should we not worry about who will buy government bonds, if not the banks, to finance the governments’ replation, liquidity and recapitalisation measures? A bank run is when depositors panic and seek to withdraw their deposited savings simultaneously from a bank, ironically the panic creates the insolvency. But, this is only one side of the matter. When multiple bank runs occur at once, widespread banking panic also millions of customers become insolvent along with the bank. When, as now, the systemic effects are impossible to control or calculate, the idea of leaving matters to ‘Chapter 11′ for many banks all at once is economic suicide. And I am amazed that such an devastatingly obvious fact needs rehearsing to financial market academics?
