Friday, November 27, 2009

Picking up nickels in front of steamrollers

When I was growing up in the trading world, high Sharpe ratio was the holy grail. People kept forgetting the possibility of "black swan" events, only recently popularized by Nassim Taleb, which can wipe out years of steady gains in one disastrous stroke. (For a fascinating interview of Taleb by the famous Malcolm Gladwell, see this old New Yorker article. It includes a contrast with Victor Niederhoffer's trading style, plus a rare close-up view of the painful daily operations of Taleb's hedge fund.)

Now, however, the pendulum seems to have swung a little too far in the other direction. Whenever I mention a high Sharpe-ratio strategy to some experienced investor, I am often confronted with dark musings of "picking up nickels in front of steamrollers", as if all high Sharpe-ratio strategies consist of shorting out-of-the-money call options.

But many high Sharpe-ratio strategies are not akin to shorting out-of-the-money calls. My favorite example is that of short-term mean-reverting strategies. These strategies not only provide consistent small gains under normal market conditions, but in contrast to shorting calls, they make out-size gains especially when disasters struck. Indeed, they give us the best of both worlds. (Proof? Just backtest any short-term mean-reverting strategies over 2008 data.) How can that be?

There are multiple reasons why short-term mean-reverting strategies have such delightful properties:
  1. Typically, we enter into positions only after the disaster has struck, not before.
  2. If you believe a certain market is mean-reverting, and your strategy buy low and sell high, then of course you will make much more money when the market is abnormally depressed.
  3. Even in the rare occasion when the market does not mean-revert after a disaster, the market is unlikely to go down much further during the short time period when we are holding the position.
"Short-term" is indeed the key to the success of these strategies. In contrast to the LTCM debacle, where they would keep piling on to a losing position day after day hoping it would mean-revert some day, short-term traders liquidate their positions at the end of a fixed time period, whether they win or lose. This greatly limits the possibility of ruin and leaves our equity intact to fight another day in the statistical game.

So, call me old-fashioned, but I still love high Sharpe-ratio strategies.

25 comments:

Anonymous said...

I agree with you. I think the key is the time limit. LTCM was betting on that the market will converge eventually. Well, when is eventually?

For those who dismiss high Sharpe-ratio strategies, what are the alternatives?

- Anon

bill_080 said...

Isn't that "picking up nickels after a steamroller"?


Bill

Ernie Chan said...

Bill,
Exactly!
Ernie

Jez Liberty said...

The problem with mean-reverting strategies is that it works.. until it does not work any more. And even with discipline of cutting your positions you might end up with many small losses ultimately wiping out your equity

Given the non-normality of price distributions, it appears logical that trend following is the best system to profit from that fact (see this article for illustration).

Of course you will not get incredible Sharpe ratio but I would argue that is far from being the most important metrics (although it is the main one used by the industry...).

PS: I do however appreciate the point you make about how everybody is taking an opposite view now - which is interesting from a contrarian point of view

Anonymous said...

If not mistaken, large portion return of this strategy seemed disappeared last couple of year, comparing before 2003. Any thoughts why?

rb101 said...

Curious as to your sharpe target levels- they change with duration. What do you consider a good 1yr vs. 6 month sharpe ratio to be?

Also

Ernie Chan said...

Jez,
Any trading strategy, reversal or trending, has the problem that it works until it doesn't! In fact, from my experience, trending systems have generally shorter shelf life than mean-reverting strategies, for reasons I detailed in chapter 7 of my book.
Ernie

Ernie Chan said...

Anonymous,
There are many mean-reverting strategies in various markets, and I don't think that their returns have all disappeared since 2003. In fact, I have found (as did many other traders) that 2008 to be an extremely good year for short term MR strategies.
Ernie

Ernie Chan said...

rb101:
I consider strategies with Sharpe of about 2 to be good.
Ernie

Anonymous said...

2008 was a golden year for mean reversion strategies. Long term backtesting will easily show that.
However, looking at 2009 returns, it is clear that these strategies are losing steam rapidly. No surprise, given their current popularity.
Time to go back to trend following?
EB

Anonymous said...

Ernie,

Any hypothesis why this strategy exits? Is it effectively a long volatility strategy, provided it worked extremely well in 2008, or should it be profitable regardless what type market condition we are in?
Do you believe that "these strategies are losing steam rapidly" as increasing popularity?

FAN

Ernie Chan said...

Fan,
Certainly mean-reverting strategies long volatility. But more fundamentally they provide liquidity to the market, and as such, is likely to be enduring and profitable in most market conditions. I don't believe that these strategies "lost steam", though of course the returns are lower than 2008 due to reduced volatility.
Ernie

Peter said...

The real problem with the Sharpe ratio as a measure is that it is highly time dependent. I prefer to plot the Sharpe ratio through time, or use another measure of efficacy.

Peter

scheng1 said...

High Sharpe Ratio?! You must be really old!
The lecturer for our CFA tutorial used less than 5 minutes on Sharpe Ratio

Anonymous said...

Late to the game here (just found this post), but aren't you missing the main benefit of short term strategies, which is that, approximately, sd(sum(returns)) scales with the square root of the number of bets whereas sum(returns) scales linearly - and since we generally care about annual sharpes - sharpe will be approximately something like sqrt(n) * E(r) / E((r - E(r))^2) where r is the return per individual trade and n is the number of trades.

While my math is half-assed and approximate (and probably wrong), the main gain is that since we care about returns on the same time scale no matter what strategy we are talking about, short term ones give many more bets and thus a higher probability of realizing the mean (your "edge") in a fixed time scale vs. a long term strategy which will give you a small number of draws from a distribution (and thus more variance).

cordura21 said...

Hey guys. Speaking of Sharpe Ratios, did you read
Faber's thread about the inexistence of >1 reported Sharpes

I wonder if it's just an issue of scalability. Cheers, Cord

bill_080 said...

cordura21,

I wouldn't be surprised if scale becomes a limiting factor.

For me, on a small scale, trading costs (spread + fees) are my biggest constraint. If my trades were big enough to consistently push the market around, that slippage could easily become "the" constraint.


Bill

Ernie Chan said...

Anonymous,
Yes, the larger number of bets we can make in short-term strategies is one of the main reasons I like them.
Ernie

Ernie Chan said...

cordura21,
The research you cite is certainly amazing. But from what I understand, CTA's typically trade trending strategies, which are inherently more unstable than short-term mean-reverting strategies. If the study include all hedge funds (including equities funds and multi-strategy funds such as Renaissance Technologies which I am not sure registered as a CTA) , I think there will be quite a few which has Sharpe > 1 for over 15 years.
Ernie

Anonymous said...

Ernie,
I think you are right. Most CTAs do trade trend-following strategies. Which makes me think: If short-term mean-reverting strategies have a higher Sharpe ratio, why aren't there many more CTAs trading this way? Is it much harder to discover good short-term mean-reverting strategies than trend-following ones? After all, some CTAs have substantial research departments. They SHOULD be able to find something. Your thoughts? - Zeke

Ernie Chan said...

Zeke,
I am not sure that futures markets mean-revert very much!
Ernie

Stan Maydan said...

Ernie, in order to calculate Sharpe Ratio for mean reverting strategies, I think we need to have a consistent way to calculate returns for portfolios with short positions, which are needed for mean reversion.
This is what I would like to ask your opinion about.
The first case is a static portfolio. Specifically, I would calculate long return and short return separately, but then how do you calculate weights of each?
For example positon longPosition is 100 at time t1 and becomes 110 at time t2. On another hand, shortPosition is -50 at time t1 and become -40 at time t2. In my opinion I would first calculate long return as 110/100 = 1.1 and short return as 50/40 = 1.2 and then I would weight them in the following way to calculate the total return:
(1/(100 + 40 ))(100*1.1 + 40*1.2) = 160/140 = 1.14.
Do you agree with this approach? Also, how would you modify this approach if you need to rehedge your short position between t1 and t2. For example, you need to go more short (-5) at time point T1a> t1, and then at time point T2a < t2 you need to buy back 6?
Thank you

Ernie Chan said...

Steve,
1) Actually, there is no requirement that we must have short positions to have mean-reversion.
2) Daily return on gross market value of your portfolio is very simply calculated as P&L/(long market value + abs(short market value)). No need to worry about weights. This is the return that should go into Sharpe ratio calculation.
Best,
Ernie

Stan Maydan said...

Ernie, thank you for your answer. How do you calculate drawdown in the portfolio with short positions? How do you adjust drawdowns if you need to rehedge your short position (by going more short or buying back), as in my previous example?
For example, your original portfolio at t1 is 1,000 long of S1 and 250 short of S2. Assume that at t2 it is 900 long and 220 short. At this time you short 100 more of S2. Now at time t3 your long is 880 and short is -300. Your long drawdown is 120. How do you calculate your short drawdown? How do you calculate your total drawdown?

Thank you

Ernie Chan said...

Stan,
First, daily return =dr=dailyPL/gross_mkt_value.

Then, you can calculate cumulative return by the usual (1+dr1)(1+dr2)...-1.

Lastly, drawdown is the cumulative return starting from the high water mark.

You do not need to calculate the long and short side separately.
Ernie