A quick reprise on the BIS study purporting to show that Frankendodd and EMIR will increase economic growth by better than .1 percent per year.
A footnote in my earlier post questioned whether multilateral netting via single-category CCPs really reduced exposures substantially, if at all, compared to cross-product bilateral netting. The BIS study assumes-without any support that I can identify-that multilateral netting via CCPs is four times (!) as effective in reducing exposures as bilateral netting.
Again, I’m aware of no theoretical or empirical support for this number. The standard Duffie-Zhou theoretical analysis doesn’t reasonably support such a conclusion, and that is a purely theoretical exercise. Such a remarkable assertion should require some empirical support.
More importantly, this assumption begs a huge question: since derivatives counterparties internalize the benefits of netting (even when it imposes external costs on other claimants), if multilateral netting compresses exposures by so much, why weren’t banks falling over themselves years ago to move trades to CCPs, thereby reducing (shifting, actually) risk and freeing up capital? Why is it necessary to unleash Frankendodd to force dealer banks to pick up this money off the sidewalk? I thought they were greedy SOBs.
I see only two answers. First, that the netting economies really aren’t that large. Second, the dealer banks recognized that that increased netting merely transferred exposures to other creditors, who would adjust the prices of credit accordingly, making the (alleged) reduction in derivatives exposures not worth the cost of setting up and operating a CCP, posting collateral, etc.
Either way, there is no way to reconcile the BIS report’s claims that (a) netting of derivatives exposures is true net benefit, and (b) this net benefit can be achieved only as a result of regulation that forces dealer banks to clear to achieve multilateral netting economies. Indeed, to the extent that netting redistributes wealth from non-derivatives creditors to derivatives counterparties, dealer banks would have an excessive incentive to adopt clearing/multilateral netting.
In other words, the BIS report implicitly assumes that greedy bankers stupidly overlook oodles of money ready to be stuffed into their pockets.
Why do I get the impression that this was a put-up job intended to give the G20 something to crow about at the upcoming meeting in St. Petersburg?
John Kiff of the IMF tells me via Twitter (where he tweets as @kiffmeister) that the report recognizes some of the problems I identified in my post, namely that there are other sources of systemic risk than the counterparty risk channel that is the basis for the analysis. Yeah, that’s true, but I have looked through the report carefully, and can find no recognition whatsoever of the fact-and it is a fact-that collateral and netting have redistributive effects.