Another Bad Idea, Courtesy of Dodd-Frank

Dodd-Frank says banks should stop behaving like hedge funds, which means, in practice, that they should stop making "bets" with their capital and simply make loans and service customer transaction flow. Like many ideas from Washington, this sounds great to anyone not actually acquainted with the situation.
For starters, how can you tell? Any trader has a "book" which shows what they are currently long or short. Is this speculation or customer-driven trading?

Allow me to explain. Let's say I trade treasuries notes. I can't always have a zeroed-out book. If someone want a bid on, say,  $10 million 10-years, I have to give them one. If the customer says, "done," I own them. Now, if I immediately sell them to someone else I assume the Fed will not think I am speculating. But what if I can't? What if I don't want to because I think the market will tick up this afternoon? What if I want to just carry them for customers who might (or might not) be looking for offers later?

The point is, where is the line between normal transaction business and speculation is impossible to define, and yet that is exactly what the Fed proposes to do. Their answer on how do accomplish this is to use the Sharpe Ratio.

For those of you not in finance, the Sharpe Ratio is a simple metric where you (basically) divide your returns by the volatility of those returns. High Sharpe Ratios are thought to be good because they mean you are getting good returns relative to the risk you are taking. (This assumes, of course, that volatility equals risk.)

The Fed would look at the numbers from each trading unit and say, "You are not speculating because you have a high Sharpe Ratio. Good boys." Or the opposite.

I have a number of problems with this idea.

First, high Sharpe Ratios don't necessarily suggest low risk.  In fact, really high ratios actually mean the opposite: run away. Long Term Capital had an excellent Sharpe, for example.

Without going into excessive detail, high Sharpe strategies tend to fall into two categories. The first is arbitrage strategies (like Long Term) where one takes small pricing inefficiencies and exploits them with a lot of leverage. The other is illiquid lending strategies (commonly known as PIPES) where one loans money to micro cap companies that can't otherwise raise capital easily. Both strategies look great right up until they don't.

Second, the need to show a high Sharpe will drive all the trading desks to similar strategies, thereby increasing systemic risk. If everybody holds the same positions, what happens when they all run to the exit at once? This is guaranteed to happen.

Third, let's say the Fed decides your desk's Sharpe is too low. Then what? Do they order you to do something? If so, what? Does the Fed know better than an actual trader how to reduce risk? (The answer is no, in case you're dwelling on this.)

There is a better way, and a simpler one. What got us into the Big Mess of 2008 was, mostly, leverage. Too much of it, to be precise. Placing reasonable institutional-level limits on leverage make sense, and it doesn't require anyone to make qualitative judgments they're not equipped to make.

But leave it to Washington to find a more complicated way.