The Fed Will Take 174 Pages To Tell You What "Prop" Trading Is


Paul Volcker has made himself surprisingly relevant recently as an enemy of both (1) inflation/whatever Paul Krugman might be up to and (2) proprietary trading/whatever Kweku Adoboli might be up to. As for the second category, on Friday I guessed that I was not alone in being confused by the Volcker rule, which would ban "proprietary" trading by big banks while still allowing not-"proprietary" ("flow"? "customer facilitation"? "market making"?) trading. So I was pleased to learn today that the regulators designing the Volcker rule seem to be equally confused:

According to a 174-page draft of the rules seen by the Financial Times, and confirmed by people familiar with discussions between regulatory agencies, so-called “repo” transactions and securities lending, and near-term trading in currency and commodities – but not futures – will be permitted.

The draft rules exempt from the prop trading ban “positions arising under certain repurchase and reverse repurchase agreements or securities lending transactions [and] bona fide liquidity management”. They also allow “positions in loans, spot foreign exchange or commodities”.

So ... what? Sadly the FT is just picking passages from a 174-page draft rule, and I assume this rather misrepresents the intent. An asset class is unlikely to be "proprietary" or not as a whole. Positions in foreign exchange can be proprietary or customer facilitation. Positions in loans can be proprietary trading positions, customer-facilitation trading positions, or ... what is it called when a bank lends money to a client? (That's its own capital right?)

Or maybe it's this:

The draft, dated last month, would ban short-term trading that might “significantly increase the likelihood that the banking entity would incur a substantial financial loss or would fail”.

Good thought: "don't do trades that have a high probability of blowing you up." But, also, not necessarily a clear bright line.

I can think of four ways to implement a rule that says "you can facilitate customer trades with your own capital, but you can't trade on your own account with your own capital." First, you could do a really formalist approach - you say that any desk called a "prop" desk has to shut down, but you give dealers wide latitude to figure out how much risk they're willing to carry on their flow trading desks and how they want to hedge it. That approach has the lowest possibility of doing active harm - of banning trading activities that are Good for the World - but it also has the least likelihood of solving the problems the Volcker rule is intended to solve. Because if a bank wants to take on a risk, it has a lot of flexibility to do so as an adjunct to its customer business.

Second, you could have a principles-based approach: you could, say, "ban short-term trading that might 'significantly increase the likelihood that the banking entity would incur a substantial financial loss or would fail,'" or just generally ban things that seem proprietary, and maybe (as the FT says has been considered) require bank CEOs to certify that they do not engage in proprietary trading without working too hard to define in advance what "proprietary" means. This makes a kind of sense - many people on Wall Street have a sense of what they think is "proprietary" and what they think is "customer facilitation" and related hedging, so they can manage on principles. But it relies on everyone sharing the same approach - and there's a decent chance that regulators and traders/managers will disagree. Worse, there's a chance that juries will disagree - that CEOs who certify that they're not doing any prop trading will find out, once they've been arrested, that what they think of as customer-related risk management sure looks in hindsight like "short-term trading that increased the likelihood of a substantial financial loss." It's hard to run a company based on loosely defined terms when the penalty for getting it wrong can be jail.

Third, you could have some sort of statistical approach. The FT reports that "[s]wings in a trading desk’s 'value at risk' will be one of a number of red flags used by regulators to decide whether they are engaging in risky, prohibited activity." That makes plenty of sense if you want to regulate - VaR. If your goal in having a Volcker rule is to regulate the amount of capital that banks are risking, then you can do that. To some extent, Basel III and other bank-capital rules are designed for just that purpose. If your goal in having a Volcker rule is to "ban proprietary trading," though, VaR doesn't necessarily measure that. It measures risk, not prop-ness.

The fourth approach you could take is just try to imagine every situation possible in the world and decide whether it's "prop" or "not." That seems like an obviously losing strategy, as the world is pretty good at proliferating around rules. But at 174 pages, that sounds a lot like what the regulators are up to.

Fears over exemptions to Volcker rule [FT]