My frequent sparring partner, John Kemp of Reuters, reads the IHS study justifying the benefits of bank involvement in commodity markets, and is not convinced. I agree that the report is rather lame, but am not convinced by Kemp’s critique, and the policy implications he draws.
One of Kemp’s points is that banks are not essential to the purchase, sale, transportation, storage, or processing of commodities. Other entities, most notably commodity trading firms like Vitol or Glencore or Trafigura, can provide the same services as the banks.
This is true, but beside the point. There are few types of firms that are utterly essential, with no substitutes. Even banks are not irreplaceable in the functions that Kemp, and others, agree is appropriate for them. Their core function of providing credit can be performed by the capital markets. Or loan sharks.
The fact that banks can profit as a result of their activities in physical commodities should create a rebuttable presumption that they are providing economic value. In a reasonably competitive market, where costs and benefits are internalized, firms that survive do so because they can provide a good or service of equal or better quality at equal or lower cost to other potential suppliers. Indeed, in these circumstances, all active sellers have marginal costs that are approximately equal: cost functions determine the relative outputs of the active suppliers. Driving out suppliers who can profitably service part of the market will drive up prices, reduce output, and reduce welfare: some of the output supplied by the eliminated firms will go unproduced, and the rest produced at higher cost.
Competition is not the issue. The market for commodity intermediation is highly competitive: even the big banks account for a relatively small share of activity, and the biggest players account for no more than 5-6 percent of trading activity in any major commodity market. And even if banks accounted for a big share, that could be due to a cost advantage.
Which brings me to Kemp’s second objection, which potentially has more merit. The above says “where costs and benefits are internalized.” Kemp brings up, quite properly, the possibility that banks get a subsidy as a result of too big to fail, and hence the true costs of their activities are not internalized.
The immediate implication of this is that banks are too big, and that there market share in physical commodity activities is too big. But “too big” does not mean that the right share is zero. Banning bank participation in physical commodity markets, or excluding them from certain activities, is likely to be an overreaction, not a Goldilocks (“just right”) strategy. They might be too big now, but zero is almost certainly too small.
Note too that eliminating banks from physical commodity markets, or restricting their participation, does not eliminate the TBTF subsidy. Yes, they might not be able to exploit the subsidy by getting too big in commodities, but they are not going to let the subsidy money burn a hole in their collective pockets. Won’t let them exploit the subsidy in commodities? They’ll find someplace else to do it. Restricting activity in one market segment will not appreciably reduce the TBTF subsidy that banks reap.
The right way to address this problem is to find ways to get banks to internalize the true costs. This is best done by capital requirements that reflect the risks of the activities banks engage in, and the costs associated with bail outs. Restrictions on permissible activities are a meat cleaver approach that will not mitigate the TBTF subsidy.
Kemp also goes on about the metals warehouse games. As I’ve said repeatedly, both here and in media interviews, banks are not uniquely positioned to engage in such sorts of shenanigans. If the conditions are right, whoever owns the warehouses will play the games. The best way to deal with this problem is ex post deterrence, either through private action (e.g., the numerous class action lawsuits against Goldman for its alleged warehouse manipulations) or government enforcement actions.
The role of banks in physical commodity markets will remain under scrutiny-and attack-for the foreseeable future. Many of these attacks are predicated on ignorance, not to say stupidity. Kemp isn’t guilty of those things, but his analysis is nonetheless misguided. Even when he identifies a real potential problem-the TBTF subsidy-eliminating banks from the physical commodity market is not the appropriate way to address it.
This illustrates a general point I’ve mentioned in many forums. Attempting to address TBTF problems, and systemic risk problems, through structural interventions such as clearing mandates, SEF mandates, Glass-Steagall, or preventing banks from participating in commodity markets, is inefficient and likely ineffectual. It is better to operate at the level of fundamental incentives, by getting banks and other financial intermediaries to internalize the costs and benefits of their activities. This is an easy principle to state in the abstract, though hard to implement in practice: just what is the right capital charge overall, and how should different bank activities contribute to this charge? Not an easy question to answer. Although imperfect, that’s still a better way of addressing these problems than mandating market structures.