As the Dodd-Frank rule-making process begins in earnest, I’d like to offer some unsolicited advice to regulators, and specifically the Fed. I spent many years at a large dealer, so I have some thoughts on how an effective supervisory regime for large dealer banks needs to be structured.
Put simply: You need to get in the banks’ face. I’m deadly serious about this. First, significantly expand the dedicated supervisory teams for the dealer banks that qualify as Tier 1 FHCs. It’s not enough to have a 5-10 person supervisory team for dealer banks like JPMorgan, BofA-Merrill Lynch, Morgan Stanley, etc. The capital markets side of each of these banks has tens of thousands of employees, and hundreds of people in senior risk-taking positions. The supervisory team for each Tier 1 FHC needs to have at least 50 people. Again, I am deadly serious.
Second, and most importantly, at least half of the supervisory team needs to be on-site full-time. The CPC (“central point of contact,” who heads the supervisory team) also needs to be on-site, and should have broad information-gathering authority. In other words, the Fed should be a major presence at each dealer bank. Most of the on-site supervisors should be assigned to the Sales & Trading side of the bank, since that’s where most of the short- to medium-term risk is.
The reason for all this is simple: The banks’ trading books turn over constantly, and given how large these trading books have become (relative to the bank’s total consolidated balance sheet), the risk profile of a dealer bank can change very rapidly. It’s flat-out impossible to have a realistic, real-time understanding of the health of a dealer bank without being on-site every day — I’m not even willing to debate this point.
Supervisors need to be at the bank’s main offices — conducting rotating examinations of different business lines, independently monitoring compliance with things like the Volcker Rule, as well as simply collecting anecdotal information, both about the markets and the firm. It’s easy to dismiss anecdotal information as unreliable, but this kind of color can be incredibly useful, especially for regulators. If you only rely on fully-vetted, reportable information, and reports prepared for the board of directors, then you’re guaranteed to be hopelessly behind the curve. You’re far more likely to identify potential problems at a dealer bank, as well as potential sources of systemic risk in the markets, by simply talking to people at the bank regularly.
Traditionally, supervisors for LCBOs (large complex banking organizations) have focused a disproportionate share of their efforts on ensuring that the bank has adequate risk management systems and internal controls. Given supervisors’ limited resources in the past, I actually understand why they chose this approach. But this cannot be supervisors’ primary focus anymore. No matter how sophisticated or expensive their risk management system is, banks can always just ignore internal limits. When a profitable desk uses up all its balance sheet allocation too early in the quarter, or comes up against its risk limits, there’s a good chance the desk will be granted a waiver, or simply have its risk limits raised (“just this once,” they’ll be told, although somehow it never works out that way). Even if the bank tells its supervisors about the breached risk limits — which isn’t a given, by any means — the deed is already done. No, supervisors need to take some responsibility for monitoring compliance, and they need to do it on-site.
Now, I know what the banks’ response would be to this proposal. They would vehemently object to the higher fees they’d have to pay to the Fed to fund this supervisory expansion. They would also be particularly upset about having to give Fed supervisors office space in their main New York offices. (“Do you know how valuable this space is? We have vice presidents working in cubicles!”) And they would complain about the increased compliance costs that would be required to provide the on-site examiners with the information they need (which they would no doubt supplement with a wildly inflated estimate of said compliance costs).
Please, I beg of you, do not give these arguments the time of day. The proper response to these arguments is:
“Yeah, well, life’s tough in the aluminum siding business. Deal with it. This is the cost of being a large dealer bank with access to the Fed window. As long as you have bonus pools in the $15-20 billion range, you’re not allowed to complain about the increased fees. If you don’t like it, you should consider another line of work.”Think about it.