The 10th Man

Keep Your Risk Managers Away From Me

January 22, 2015

This week’s big news, of course, continues to be the massive revaluation of the Swiss franc (CHF). It’s perhaps the first instance of a G10 currency going up 16% in a single day.

From a strategy standpoint, there really is only one way to interpret this, as many people already have: it’s the end of central bank omnipotence.

Central bank says it’s going to do A, does B instead. For investors, it’s much harder to take risk in that kind of environment. So I think the logical thing to do is to look at other pegged/managed currency pairs in the world—like the Chinese yuan, the Hong Kong dollar, and the Danish krone—but also any situation where a central bank has said it’s going to do an unlimited amount of anything, because as you can see with the Swiss National Bank (SNB), it’s subjected to the same P&L forces as everyone else.

This $5 Trillion Market Is Just Getting Started.

Don’t miss out on the ETF revolution. Get going with this must-read report from Jared Dillian.

Moving right along, I want to talk about the risk management aspect of this trade.

Within a few hours, we knew that a couple of retail currency brokers needed capital, or else. And we learned that Polish and Hungarian folks who took out CHF-denominated mortgages were in big trouble too. But more important, anyone who was just plain old short the Swiss franc was also hosed with or without a stop loss, which wouldn’t have made much difference in this case anyway.

I’ve been trading for 15-plus years, and I have never blown myself up. (If I had, I probably would not be writing this.) Don’t get me wrong, I’ve done plenty of dumb things over the years, made lots of bad, even stupid trades, and I have occasionally let my losses run a bit too long. But I have never been hit by a Mack truck—walking into work and suddenly finding myself suffering catastrophic losses.

Knock on wood.

How Risk Managers Can Get You in Trouble

Let’s discuss the margin system that most FX trading shops use (and I use) for a moment. There are a lot of dumb journalists running around saying, “Why the hell are FX brokers offering their clients 25 or 50-to-1 leverage on currencies? More regulation!”

Well, this isn’t anything new, and the margin system works well most of the time (note that this is basically the same system used by derivatives exchanges, on which everyone wanted to list credit default swaps at one point, because exchanges don’t fail, remember?).

This time it didn’t work so well. The margin calculators looked at euro/Swiss franc (EURCHF) volatility, which was very low, and afforded their clients high leverage based on that low volatility. They didn’t take into account the fact that EURCHF was right in the neighborhood of the floor and that if the Swiss National Bank abandoned the floor, there would be a ridiculously large move.

This is one of the reasons that risk managers (i.e., non-traders in these positions) can mean trouble. An experienced trader looks at EURCHF at 1.205 and says, “Nope, not touching that.”

It has nothing to do with his opinion on the Swiss franc being undervalued but everything to do with the fact that the currency pair is near a self-imposed floor, and if the floor is removed, the risk is very asymmetric. A risk manager may calculate the volatility of EURCHF and find it negligible, but the trader looks at EURCHF and says, “This pair might have 2 volatility to the upside, but 80 vol to the downside.”

So it’s not the margin system itself that failed—it’s the idiots setting the margin levels who got it wrong. And even then, you still cannot plan for a black swan. Even if margin had been 10-to-1 instead of 50-to-1, people still would have gotten rinsed.

This is me editorializing: I think the scary thing about present-day markets is that you have fewer experienced traders in charge. In an age where 50% of stock market volume is algorithmic, it’s young folks with math and physics degrees who are building these models and who have no sophisticated understanding of actual risk, the kind that cannot be mathematically modeled.

They probably think they’re superior to people like me because they know math and I don’t, but when you’ve traded two of the biggest bear markets in the last century, you begin to learn that the market has a much bigger (and more malignant) imagination than you.

It’s funny—many traders are good at protecting against one specific kind of tail risk (the hypothetical terrorist attack), where they buy S&P puts and VIX calls. Yes, it’s possible that we walk in one day and the entire market is down 20%. And one thing you learn from being a floor trader is that nobody sleeps well at night, especially if you’re short volatility, because what if?

But tail risk exists all over the place, and I think the Swiss franc incident was a pretty good reminder.

I remember back in 2000-2002, you had companies that would gap up 50% (PaineWebber slapping a $1,000 price target on Qualcomm shares comes to mind) or down (Enron, WorldCom, etc.). You could not have any unbounded risk at all.

This $5 Trillion Market Is Just Getting Started.

Don’t miss out on the ETF revolution. Get going with The 5 ETF Trading Strategies You Should Know About Before Investing, from Jared Dillian.

Nowadays, people have gotten more comfortable with unbounded risk. I pointed out a few issues ago that the volatility of volatility (vol of vol) was going up. It is still going up. Oil just went down 60% in a matter of months. Another black swan. One of the reasons I got to where I am today is because I was one of those guys who would buy back the nickel puts before expiration. It’s a good habit to get into.

Jared Dillian
Jared Dillian

 

Get Thought-Provoking Contrarian
Insights from Jared Dillian

Discuss This

0 comments

We welcome your comments. Please comply with our Community Rules.

Comments

Dallas Kennedy

Jan. 22, 2015, 4:55 p.m.

A great article, and I’m a physicist myself. Having watched the “efficient, slowly changing, always-continuous, always-liquid market hypothesis” follies of the last 20 years, including watching friends swimming in them, I have to say: you can model this stuff, but it’s hard.

It has to be done with functions that allow near-singular behavior, or low-statistics simulations that allow for one or a few actors to have a disproportionate impact. These phenomena underlie the so-called “black swan” or “fat-tail” statistics called “Levy-stable distributions” by mathematicians. They’re the residue of chaos (in the technical sense). Such distributions are one way to generalize the bell-curve (Gaussian) distribution.

But it’s hard, as I said: you don’t have a lot of empirical data to work with, the functions in practice are highly sensitive to small changes in assumptions, and there are no nifty, compact formulas. That last point is a big problem with attempts at standardized risk management, which requires reproducible cookbook formulations.

That said, you can do something, which is to work with these distributions indirectly, using correlations and moments. You can analyze both the time-based periodic components and the unique, one-off events using something called “wavelets” (related to Fourier analysis, which by itself doesn’t work here). There are other methods. But they’re not in undergraduate textbooks.

We need more academic research into how financial systems actually work. Academic economists are typically not interested in finance. People trained in advanced finance usually go to work in the industry and, however good they are, usually lack the time and detachment to be objective observers. We need disinterested scientific objectivity here. Andy Lo at MIT is someone who springs to mind.