Posts this month
A blog on financial markets and their regulation
Gary Gorton and Andrew Metrick published a fantastic paper last month on “Securitization” (NBER Working Paper 18611). This paper contains a wealth of information, a detailed survey of the literature and a number of very interesting theoretical ideas. What I found most interesting is the idea that the most important benefit of securitization could be a reduction in bankruptcy costs. In passing, Gorton and Metrick talk about “contractual living wills” a set of contractual arrangements in securitization that have some similarities to the living wills that are being proposed as mechanisms to enable easy resolution of banks in the post crisis regulatory reforms. I think this analogy is worth pursuing even further.
In a securitization, all the assets and liabilities are housed in a Special Purpose Vehicle (SPV) which is structured in such a way as to make bankruptcy all but impossible. Gorton and Metrick see this as a big part of the economic function of securitization:
… the SPV cannot become bankrupt. This was an innovation. That is, the design of SPVs to have this feature is an important part of the value of securitization. Moreover, it has economic substance. Since the cash flows are passive, there are no valuable control rights over corporate assets to be contested in a bankruptcy process. Thus, it is in all claimants’ interest to avoid a costly bankruptcy process. (Page 19)
If the assets perform badly and the cash flows from the assets are not sufficient to pay all the coupons, the SPV does not enter bankruptcy – instead the available funds are used to pay the senior claimants early while writing down the liabilities to the junior claimants. Gorton and Metrick call this a contractual living will (Page 8). But I think it is much more than the living wills that post crisis banks are being required to prepare for themselves. It is not just that the SPV waterfall rules are contractual and therefore self implementing unlike the wishful thinking that goes into the living wills of the banks. What is more important is that the SPV waterfall rules constitute a contractual bail-in arrangement whereby the junior claimants’ principal gets written down to restore the solvency of the SPV. Similarly liquidity problems are automatically addressed by extending maturities contractually. (It is not uncommon to see securitization structures in which the expected weighted average life of a securitization tranche is only 5 years, but its rated and legal final maturity is 30 years.)
Gorton and Metrick are right to point out that some of these things are easy to do because the cash flows of an SPV are passive and therefore there is no judgement required to manage them. The SPV is “brain dead” and is completely governed by contract. But I think that resolution of banks and other financial institutions can learn a lot from the SPV liquidation arrangements. Failed institutions can often be put in run-off mode where most of the management can be passive. Private ordering usually fares better than complex regulatory mechanisms.
It is also possible for a business segment to be put into SPV style liquidation arrangements (with near zero bankruptcy costs) while the rest of the institution runs normally. Many central counterparties (CCPs or clearing corporations) have framed rules under which if the losses in a particular segment exceeds a certain threshold, then loss allocation mechanisms kick in that would effectively shut down that segment – contractual bail-in eliminates bankruptcy. I think regulators should consider mandating such contractual provisions that make it impossible for a CCP to go bankrupt. CCPs should be allowed to fail, but the failure should not involve bankruptcy. Post crisis, many CCPs are beginning to clear very risky products that make it extremely likely that a large CCP in a G-7 country would fail in the next decade or so. Contractual living wills and contractual bail-ins would prevent such a failure from being a catastrophic event.
I think it is also possible to convert a failed bank into a CDO that is put into run-off mode with contractual provisions governing the loss allocations without any need for formal bankruptcy at all. Nearly seven years ago (well before the global crisis), I wrote in a blog post that “Having invented banks first, humanity found it necessary to invent CDOs because they are far more efficient and transparent ways of bundling and trading credit risk. Had we invented CDOs first, would we have ever found it necessary to invent banks?” Even if we do not want to replace all banks by CDOs, we can at least replace failed banks by CDOs that are “liquidation efficient” in Gorton and Metrick’s elegant phrase.