Seven Habits

Martin Wolf's blog has some suggestions for improving financial regulation. As usual, quote plus comment:

First, coverage. Perhaps the most obvious lesson is the dangers of regulatory arbitrage: if the rules required certain capital requirements, institutions shifted activities into off-balance-sheet vehicles; if rules operated restrictively in one jurisdiction, activities were shifted elsewhere; and if certain institutions were more tightly regulated, then activities shifted to others. Regulatory coverage must be complete. All leveraged institutions above a certain size must be inside the net.

OK, but this is a colossal task and one laden with systemic risk. It covers the regulation of banks, broker dealers, hedge funds and insurance companies globally. Do we really think it is realistic to work towards common regulatory standards for all of these? And if we do, we run enormous regulatory risk: set the wrong standards for *everyone* and *everyone* will do the wrong thing. At least at the moment regulatory diversity ensures that different players have different incentive structures.

Second, cushions. Equity capital is the most important cushion in the financial system. Also helpful is subordinated debt.

I might quibble with that last sentence to some degree but certainly leveraged players need capital. This is motherhood and apple pie.

Third, commitment. The originate-and-distribute model has, it is now clear, a huge drawback: originators do not care sufficiently about the quality of loans they plan to offload on to others. They do not, in Warren Buffett’s phrase, have “skin in the game”. That makes for sloppy, if not irresponsible or even fraudulent lending. Originators should be required, therefore, to hold equity portions of securitised loans.

Or at least some of the equity portion. And, in order to stop people selling very thin equity tranches then leaving the real risk in the mezz, 'equity tranche' should be defined to at least include the capital required for the whole tower calculated at a reasonable confidence interval and with conservative default correlation assumptions.

Fourth, cyclicality. Existing rules are pro-cyclical. Capital evaporates in bad times, as a result of write-offs, thereby forcing contraction of lending, worsening the economic slowdown and further impairing assets. Mark-to-market accounting, though inherently desirable, has a similar effect.

Agreed. We need a mechanism to make net capital requirements anti-cyclical, and to counteract the unpleasantly procyclical combination of constant leverage and mark to market.

Fifth, clarity. Lack of information, asymmetric information and uncertainty are inherent in financial activities.

I doubt the efficacy of information simply because I believe in the power of lazyness. The real contents of subprime ABS were not hard to discover. Investors didn't buy on a rating because they couldn't do their own due diligence: they bought on a rating because they couldn't be bothered to do their own due diligence. More disclosure is good but I don't expect it to make much difference.

Sixth, complexity. Excessive complexity is a significant source of lack of clarity. It is particularly damaging, as we have seen, to the originate-and-distribute model, because markets in complex securitised products may, at times, seize up, forcing central banks to become “market makers of last resort”, with all the difficulties this entails. One possibility then is to insist that all derivatives be traded on exchanges.

Arrant nonsense. Complexity is not necessarily bad and neither do exchanges necessarily provide either price transparency or a guarantee of liquidity.

Seventh, compensation. On this I can do no better than quote Mr Volcker: “In the name of properly aligning incentives, there are enormous rewards for successful trades and for loan originators. The mantra of aligning incentives seems to be lost in the failure to impose symmetrical losses – or frequently any loss at all – when failures ensue.” Whether regulators can do anything effective is unclear. That this is a challenge is not.

The incentives in favour of the first institution to figure out how to break the spirit of any regulation here are so huge that I doubt we will see at best short term and ineffectual change.