Vikram Pandit has a clever idea:
It is not enough to require financial institutions to disclose capital ratios. Without knowing what that institution’s underlying assets are (only insiders and select regulators know that), outsiders, including most investors, cannot properly assess how that institution calibrates risk.
What is needed is a way to compare apples with apples. Regulators should create a “benchmark” portfolio and require all financial institutions, not just banks, to measure risk against that. The benchmark portfolio would not actually exist on the balance sheet of any one institution. Rather, it would be a collection of real investments that stand in for the kinds of assets that most financial institutions actually hold at the time. What is more, its contents would be 100 per cent public.
Institutions would be required to produce, on a quarterly basis for that benchmark portfolio, a hypothetical loan/loss reserve level, value at risk, stress-test results and risk-weighted assets. Right now these measures are run only against an institution’s actual portfolio and only a limited number of the results are disclosed. Worse, those results have no common frame of reference. The benchmark portfolio would supply that needed frame of reference.
As Lisa Pollack and Kate Mackenzie say, the problem here is that investors have to be able to trust that banks are modeling the benchmark portfolio in exactly the same way, using exactly the same tools, as they use for their own real assets. And this I think is where this whole idea falls down.
In the real world, banks know when they’re doing something risky, and then they create structures which make that risk invisible. They might, for instance, classify assets as “Level 3″ in an attempt to stop having to mark them to an inconvenient market. Or they might do what Pandit’s employer, Citigroup (C), did, and create all manner of off-balance-sheet vehicles which don’t report to shareholders or investors at all. Or they might do something else entirely which we won’t discover until after the next crisis.
That said, there’s the germ of something useful in Pandit’s proposal. So here’s how I’d improve it. Rather than having just one benchmark portfolio, or even four, as Mark Carney has suggested, give banks a kind of pop quiz every so often. One day, at about 10am, all the big banks in America would suddenly get given a surprise portfolio as selected by the New York Fed, and told “reserve against this”. With results due at, say, 3pm that afternoon. The New York Fed would then publish every bank’s answers, along with the amount of time that bank took to generate them.
By all means keep Pandit’s benchmark, too. But don’t allow the banks to game the benchmark just because they know exactly what it contains: it’s important to see how banks’ risk-management systems treat a portfolio which they haven’t been tweaked to expect.
What I’m thinking about, here, is the series of statements coming from John Thain, when he was the CEO of Merrill Lynch, saying that this latest write-down was the last ever write-down he would ever need to make against the bank’s mortgage-related assets. In that kind of atmosphere, the New York Fed could easily have published a portfolio of mortgage bonds, and asked every big bank to mark it to market and say how risky they thought it was. And that, in turn, would give a pretty good idea of just how conservative someone like Thain was really being.
And more generally, just being able to see the range of results, for any given portfolio, will be a healthy reminder that banks measure risk in very different ways. Pandit’s proposal wouldn’t just show which banks were particularly conservative; it would also show how much variation there is in the banking system as a whole. Which is in many ways even more important.
So let’s keep this concept alive, and try to make it as good as we can. As Mark Carney says, it can’t hurt. And it might just do some good.