Inherent weakness of the Sharpe Ratio

P

ProfitTakgFool

As you probably know, the Sharpe Ratio is a measure of risk for a given investment. The higher the ratio, theoretically, the lower the risk. The Sharpe Ratio has a huge flaw. If you have an exceptionally good month or two - relative to previous months - because volatility has gone through the roof your Sharpe Ratio will decline significantly and the standard deviation between returns will increase significantly. Over the past two months I have watched a full point come off my Sharpe Ratio because I've had a terrific couple of months.

The decline in this Sharpe states this investment vehicle has become more risky, not less. Interestingly enough, my drawdown over the past two months has been at it's lowest point of the year and I'm making more money with less contracts. My monthly commission rate is at an all time low and points per contract is at an all time high -- this indicates risk is low. The Sharpe states otherwise.

Moral of the story: You cannot analyze the risk of an investment by using just the Sharpe Ratio. How many hedge funds with very high Sharpe Ratios have gone out of business this year? The Sharpe has it's place but the investing community puts far too much emphasis on this number.
 
I disagree, although I don't believe in the "Sharpe Ratio" per se (or Sortini ratio, etc). What Sharpe measures is the deviations of your returns versus your equity curve. The sharpe ratio has no way of measuring the underlying volatility of the instrument (here the sp500 since you counter-trend futures); thus the sharpe ratio is measuring the deviations along your equity curve and showing that is increasing (ie your swings may be more upwardly biased, but you are swinging more around the slope than you have in the past). So, its declining because it is saying that your excess returns (everything you are showing) is more likely the result of volatility than it is alpha (excess return of your strategy). In short, even though everything looks better your excess returns are more a matter of the market than your edge, which is what the sharpe ratio is telling you.

Similarly, if volatility declines you see the inverse, lower returns, bigger drawdowns, etc but a higher Sharpe, meaning the decrease in your performance is more likely to be the underlying instrument rather than poor performance.

Ultimately, all your measuring with sharpe is deviations from your equity curve; the theory being that the closer you follow your equity curve/the less volatile your returns, the better your system/edge. Of course, this could also be used to lie since you could write a system with a very high sharpe ratio that is excludes certain historical trades that would have throw it off, depending on your belief in black swans, you might even say a high sharpe ratio is a "bad thing" or curve fit since you are backtesting again historical black swans, not future ones.

How you decide this is the subject of another thread I think; but hopefully it answers the question of why a lower sharpe ratio with higher returns your experiencing makes sense, a high sharpe ratio is not necessarily indicative of .. anything really.
 
Quote from empee:

I disagree, although I don't believe in the "Sharpe Ratio" per se (or Sortini ratio, etc). What Sharpe measures is the deviations of your returns versus your equity curve. The sharpe ratio has no way of measuring the underlying volatility of the instrument (here the sp500 since you counter-trend futures); thus the sharpe ratio is measuring the deviations along your equity curve and showing that is increasing (ie your swings may be more upwardly biased, but you are swinging more around the slope than you have in the past). So, its declining because it is saying that your excess returns (everything you are showing) is more likely the result of volatility than it is alpha (excess return of your strategy). In short, even though everything looks better your excess returns are more a matter of the market than your edge, which is what the sharpe ratio is telling you.
More...

Excellent answer.
 
Correction, I am agreeing with you but for an entirely different reason. I disagree with your reasoning about the Sharpe Ratio's ineffectiveness in your example.
 
I have never really thought much of the Sharpe Ratio. It is my understanding that it was really originally intended for the mutual fund community and somehow it spread to other investment vehicles. I completely agree w/ you PFT. The calmar sterling, and sortino ratios are better imho
 
Profit, although your higher returns has brought your Sharpe Ratio lower (due to the increased volatility in your PnL histogram), if you wait longer and continue to realize the same level of returns, you will see that the higher expected annual return pulls your Sharpe Ratio back up or even higher.

In other words, the Sharpe Ratio will eventually reflect your true performance if you continue generating higher return for several more months.
 
You must log in or register to reply here.
Back
Top