Jeffrey Rogers Hummel explains the Bernanke view of financial intermediation but raises a critical question about its policy implications.

what distinguishes banks from other financial intermediaries is not merely that deposits are used as money but also that banks, in Bernanke’s words, “specialize in making loans to small, idiosyncratic borrowers whose liabilities are too few in number to be publicly traded.” Because bank loans are especially unmarketable, a bank collapse interrupts the flow of funds more than the insolvency of other financial institutions. ..

Offsetting negative shocks to money or velocity (i.e., stabilizing M times V in the equation of exchange) with untargeted, general injections of liquidity, as consistent with Friedman’s analysis, has the added advantage of helping to clarify which banks are just illiquid and which are also insolvent, whereas direct bailouts, as implied by Bernanke’s analysis, obscures the distinction.

My main takeaway from this paper is that from an aggregate demand perspective you should never need a targeted bailout. This is obvious if you think about it. Every aggregate demand shock is either a money supply shock or a velocity shock. Any negative velocity shock can be offset by a money supply increase. Thus, if an institutional failure is going to cause a negative velocity shock, you can absorb the failure while increasing the money supply.

I think that the case for targeted bailouts has to rest on a supply-side story. That is, you think that losing Citicorp will affect the supply side of the economy in some horrible way.

I presume that Scott Sumner would endorse Hummel’s view. I am not sure how to evaluate targeted bailouts from a PSST perspective. I oppose them from a karma perspective.