SPOE: Backdoor Bailouts and Funding Fantasies?

I'm thrilled that Jay Westbrook has finally come into blogosphere with his posts on Single-Point-of-Entry.  I've blogged a little on SPOE already, but I want to highlight what I still think are two critical problems with SPOE.  In keeping with Jay's spirit, let's call them "Backdoor Bailouts" and "Funding Fantasies". 

Is SPOE a Bailout Under Another Name?

There is a fundamental tension in financial institution resolution between minimizing disruptive spillovers from firm failure and avoiding the moral hazard and taxpayer expense of government picking up the tab. To wit, if the US government were to simply guarantee all obligations of financial institutions, we wouldn't really care about bank failure. All bank debts would be equivalent to claims against the government. But the idea that the US government would explicit guarantee the entire financial system seems preposterous (except that we basically did that a few years ago).  It seems preposterous because we know it would engender enormous moral hazard and put taxpayers on the hook:  privatized gains and socialized losses.  

And here lies the real conceptual problem with SPOE. SPOE is a pretty good approach to minimizing disruptive spillovers. The only bank counterparties that would be effected by SPOE would be the investors in the structurally subordinated debt of the firm's holding company, and they will, presumably price for that risk (more about that later).  None of the other counterparties would take a haircut. All of the derivative trades, for example, would be backed by the US gov't. That's kind of nuts. 

Again, let's be really clear about this because this is the key to why SPOE would work and also the key to what's wrong with it:  in SPOE, ALL counterparties of the failed institution get paid 100 cents on the dollar except for the subdebt investors. 

One could try to remedy this by ex ante surcharges of some sort--basically a pre-funded insurance fund--but I don't think it's possible for regulators to properly price the risk ex ante on the full range of bank product lines, and all the money will flow to the underpriced areas. It's just arbitrage. There were attempts in the Johnson-Crapo GSE reform bill do to something like this and I just think it's wishful thinking. If regulators aren't capable of preventing excessively risky behavior, even when they have the tools, what makes us think they can price for it? 

Alternatively, one could try to remedy the moral hazard by having ex post haircuts (see my gavage proposal for making the bailout beneficiaries eat their own cooking), but that requires serious political stomach and will only price for losses, not actual risk. 

Is There A Market for SPOE Bank Subdebt?

There's also a key operational problem with SPOE:  there might not be any market for the holding company's structurally subordinated debt.  For SPOE to work, there has to be enough of this sub debt to absorb all of the credit losses.  But is there any market for the hold co sub debt?

I'm skeptical that there's a lot of money that's willing to purchase that sort of black swan risk on any terms, much less enough to provide the capital buffer necessary for the financial system. The hold co sub debt looks a lot like, well, the junior tranche of a CDO:  incredibly toxic, concentrated risk. There wasn't much of an organic market for those junior tranches. Instead, the market had to be created by resecuritization (sort of a self-fueled pyramid scheme). What's more, the moral hazard that is created by the implicit guarantee of most of the firm's liabilities will only increase the risk borne by the sub debt investors. There's no way for the sub debt investors to adequately price for risk because it will inevitably increase, but at an unpredictable rate. Hmm, maybe if this stuff were auction rate, with monthly auctions.  Now what could go wrong there....

What's more, there are all kinds of institutional buyers that we shouldn't let buy the subdebt because of their systemic importance. So where are we at?  Some hedge funds that buy subdebt? Is that going to be enough to fuel SPOE? 

Now, I might be wrong in my assessment of the market appetite, but here's the thing:  no one really knows if there's a market for the sub debt, and we won't until its sold. If we switch to an SPOE system, what happens if the banks can't get the subdebt fully subscribed? Do we have to go straight to resolution? There's sort of a leap of faith in the market involved here. 

This market-leap of faith is not a phenomenon isolated to SPOE. It's appeared in other contexts recently. Chapter 14 proposals blithely assume that the market will somehow be able to provide tens or hundreds of billions of dollar in DIP financing on the turn of a dime, during a financial crisis. And proponents of privatizing Fannie Mae and Freddie Mac seem to believe that there's a pool of investors ready to take on the credit risk on a $10 trillion market. In all of these cases, if the market-leap of faith is wrong, we've got a major crisis ensuing. 

Comments

[A]

Part of the problem with the TBTF punditry is that it is focused on institutions. But, as in the case with derivative counterparties, the public doesn't want big honking positions to become big honking problems. Derivatives are a distraction; when land values collapsed, the government artificially boosted the values of deposits and other endogenous currencies by promising a 1 to 1 backstop with money. TBTF is about huge exposure to rapidly changing expectations. It is useful to separate TBTF caused by poorly run institutions, and TBTF events in which poorly run institutions exist. These require separate responses.

[Doug Kelley]

Naive question, but is it simply not possible to differentiate between "real" debt (backed by physical assets) and derivatives, and basically throw the derivative counter parties over the side if the bank fails?

Surely that would help price the derivatives more accurately (and possibly reduce the level of gambling -- as opposed to investment -- in the banking system)

[Ebenezer Scrooge]

We've got to be careful to avoid confusing moral hazard with moralistic hazard.

As Professor Westbrook said, SPOE will limit the amount of derivatives a bank can underwrite, by imposing a subdebt requirement linked to the risk in the subsidiaries.

If this subdebt is reasonable in amount--on the order of once-a-century risk or thereabouts--we have left the realm of risk and are in the realm of uncertainty. Uncertainty does not create moral hazard, because nobody can respond rationally to uncertainty. If an asteroid hits the planet or California slides into the sea, nobody relies on insurance or any other market mechanisms.

Will the subdebt requirements be reasonable? I don't know--they haven't been published yet. But the issue is one of implementation--the problem is not inherent in SPOE.

[Adam Levitin]

Scroogie: Three possibilities:
(1) the subdebt requirement is too low. Moral hazard results.
(2) the subdebt requirement is too high. Inefficient market contraction.
(3) the subdebt requirement is just right (and remains dynamically so). Goldilocks nirvana ensues.

In theory #3 is possible, but in practice, I suspect it will be #1. There will be tremendous political pressure towards #1, and the banks will always argue that the requirements are actually resulting in #2 and no one wants to prevent growth. Moreover, it's hard to see requirements ever ratcheting upward, only downward, over time. The best case scenario is that this looks like FDIC premiums, but those aren't really risk-based, and they were too low for too long for political reasons. This is the same problem that is faced with trying to legislate some sort of GSE g-fee.

[Bruno Pinna]

i guess that derivatives are a distraction;
When or where land values collapsed, the government artificially boosted the values of deposits and other endogenous and other things.