Two weeks ago we published on Professor Darrell Duffie’s article “Replumbing our Financial System: Uneven Progress.” Since then we’ve given the matter some more thought and want to comment on whether US tri-party repo should be centralized at a government-backed utility work, and would this be better or worse than current market practice.
For those who missed our first installment, Duffie thinks that “tri-party repo clearing services for U.S. dealers and cash investors should probably operate through a dedicated regulated utility.” Duffie is a smart guy and clearly isn’t ready to say that this has to be done, only that he thinks it is probably a better option than the current BNYM and JPM system of clearing tri-party repo. This also doesn’t mean that Duffie is advocating for a NewBank, but potentially for a utility like FedWire. We also know (actually, everyone in the world knows) that the Fed is impatient for further reform. Our question is, in whatever form is central clearing for US tri-party repo the right thing to advocate?
We think that the answer is no: a central US tri-party clearing agent or utility is not going to create the kind of infrastructure improvement that makes it mandatory at this time. Yes, there are some clear benefits, and there are more reforms that should take place in the current system. For example, the clearing agents probably shouldn’t be in the position of providing credit to their broker-dealer clients. But looking at all the pieces, the argument for a central US tri-party repo clearing system isn’t strong enough to justify this kind of major undertaking. Here’s why:
– The market doesn’t need another Too Big to Fail institution or utility. BNYM and JPM are already too big to fail. Creating a NewBank for clearing tri-party repo, or having someone like DTCC or New York Portfolio Clearing do it, would just transfer the risk around. The FDIC’s Orderly Liquidation Authority would still be on the hook in case of a failure.
– There is an argument that BNYM and JPM are in too many other businesses and a failure in one of them could disrupt the tri-party repo market. We agree with the point but this stretches into a much broader debate, including getting into the value of returning to Glass-Steagall. Either the US markets want simple, clearly defined financial institutions that focus on one or two types of business, or they don’t. Right now they don’t, and we don’t see why repo should be taken as an isolated case.
– If the clearing banks can’t entirely ring-fence their repo clearing activities in order to prevent internal contagion, could another institution do better? Unless a NewBank is created for tri-party repo clearing, no other institution that could take it over is immune to other market disruptions. DTCC operates multiple market sectors, and New York Portfolio Clearing, while working on cross-margining across all physical, futures and repo of interest rate swaps, treasuries and DTCC-cleared repo. Could NYPC protect itself from disruption in the futures market?
– We agree with Duffie that a NewBank would be less likely to get hit by external shocks, but we think that government authorities would first want to transfer tri-party repo clearing to an existing institution. As a result, we think that the potential damage of getting hit by shocks would remain. Along the way, as Duffie notes, costs would increase.
– Europe doesn’t seem to have the same issues with the unwind/rewind process the way the US does. That bears a greater need for evaluation than putting up the idea for a new US industry utility.
Duffie has a number of other good ideas in this paper, notably the cross-regulatory supervision of CCPs that could one day open the door to a single CCP for one or more product on a global basis, thereby substantially reducing margin costs. We like that one more than working to create a new US tri-party central clearing agent. In the meanwhile, this was a good, thought provoking paper that will likely feature in the US repo reform conversation going forward.
A link to Duffie’s paper is here.