JPMC, Dimon, Hedging, and Volcker

I think that folks are missing the implications of the JPMorgan Chase (JPMC) London losses, including FDIC board member Thomas Hoenig in this Monday’s Wall Street Journal. The JPMC situation illustrates the problems that derive from the shift in “banking” from a limited form of credit intermediation, namely, commercial banking, into the general form of credit intermediation, including methods that used to be the province of investment banks. Meaning: Banks now provide credit by holding market-traded instruments that are marked to market (unlike commercial loans), and thus want (need?) to hedge exposure to minimize earnings volatility and solvency threats. This is what “London” was mostly about for JPMC.

What counts as “banking” (credit intermediation) these days can take many forms: banks can extend credit by originating and holding, by originating and distributing (bond issuances and structured finance), and by purchasing and holding debt securities originated by others. What are the implications? First, this understanding illustrates the conceptual gap in the Volcker Rule. The Rule, in its focus on market making and hedging, imagines that the bank is taking a long position for customer accommodation and needs to hedge it. That’s what a broker-dealer does, but it’s not what a bank does. A bank takes a long credit position for its own account, which it wants to (partly) hedge. Banks are always adjusting their credit position in light of their views about credit risk, in deciding who to extend credit to and their mix of assets. We want banks to do this to assure their solvency. The Volcker Rule’s focus on proprietary trading simply misses this underlying reality: extending credit and holding credit risk is a proprietary business and it’s hard to limit hedging such activity in a mechanical way. Because the hedged position is the bank’s, this hedging activity will be, at its core, “proprietary.”

Second, in hedging activities, the problem is basis risk, liquidity risk, and operational risk. JPMC illustrates all three, and shows that losses can rapidly mount even in relatively low-volatility environments in which basis risk and liquidity risk are relatively confined. (Compare Fall 2008, in which long-short strategies failed because parties needing to raise cash sold high quality assets, depressing their prices and ruining hedges.) The potential for very large losses is of systemic concern.

Policy takeaways, tentatively: First, once we get rid of standalone prop desks at banks, i.e., their hedge fund operations, what to do about hedging in connection with managing the credit asset book seems to me a morass. Dimon’s testimony is that they executed badly. I don’t think Volcker should bar this kind of activity, though regulatory review of internal monitoring is crucial. Second, that said, the nature of modern credit intermediation, which may include hedging, puts a bank at continuous risk of failure. The systemic implications are what matter here. This pushes towards three elements, which could be framed as partial substitutes: size restrictions, limits on counter-party exposure, and higher capital requirements. Hedging risk (like JPMC) is typically idiosyncratic. If a bank blows up because a hedging strategy goes badly awry, that need not sink the financial system if the bank is small enough and if interconnectivity via counterparty exposure has been sufficiently limited. Much higher capital helps too, not just to reduce the chance of single-bank insolvency, but to increase the resiliency of other banks to single-bank failure. In a strong form version of the high capital case, for financial firms we may need to reverse the deductibility of interest and non-deductibility of dividend payments to align financing incentives with systemic stability.