Sharpe Ratio and other performance statistics

Discussion in 'Professional Trading' started by BigSalad, Jan 29, 2011.

  1. BigSalad


    I have a four month track record which I'm planning to calculate some meaningful statistics on.

    First of all, with only four months there is certainly not much data for a Sharpe ratio. But, my first question is how long it would take before a Sharpe Ratio would provide anything meaningful? Of course, the more the better - but how long should I wait?

    Then, monthly I have about 100-120 roundtrips. So, already the number of trades becomes significant. But some of the trades are simultanous (or almost), and are thus obviously depending on market conditions. They're not short term, but swing trades of 2-5 trading days.

    But, from a record of now about 400 trades - what kind of meaningful statistics would you say I should calculate?

    I would like to add that the trades are spot FX, so including leverage one way or another is also of interest with regards to return calculations.

    Any input is greatly appreciated - I don't have a background in statistics although this surely is something I'd like to have :)
  2. drp7804


    Who is the intended audience of your statistics? You? Some prospective prop firm? Trying to get people to invest in your fund?

    I like to do a poorman's version of the Sharpe ratio which doesn't include the risk-free rate... just to do quick comparisons between strategy ideas, etc. Simply average return divided by standard deviation of returns. You could apply the ratio to just about any level of abstraction (eg returns per trade, per day, per week, per month, per year).

    Also max draw-down of your equity curve in terms of both depth as well as duration.

    I also like to look at the distribution of returns (again, at various levels of abstraction) to see what the lower, say, 10% of returns look like, see the overall shape of the distribution, and so forth. If you're feeling saucy, you could look at skewness, kurtosis, etc (I just use the Excel functions for that).

    I'm by no means a stats guru, but hope that helps.
  3. the1


    Be sure to include the Sortino Ratio, which, in my opinion, is a better ratio because the Sharpe punishes you for an outstanding + month.
  4. nLepwa


    For people who understand quantitative finance and basic statistics the best measure is the rolling profit factor coupled with the number of trades. With these two you can explain 100% of the equity curve.

    The investors (they don't fall in the first category usually) are usually happy with cagr and max dd. Sometimes sharpe ratio and the number of positive years/months too
    In my experience they also like to see comparison with a benchmark.

  5. BigSalad


    Thank you everyone for your input, I found all your answers very helpful. Yes, investors are my intended audience and - again, yes - they're not very quantitative minded :)
  6. tyrian


    There are several methods to calculate the downside risk in the Sortino Ratio, so we have to be careful in using consistent definitions when compariing investment results. The most common appears to be the 2nd order lower partial moment, as in this Excel spreadsheet:
  7. tyrian


    The Sortino Ratio Excel spreadsheet has moved here:
  8. A four month track record doesn't mean anything.

    A sharpe ratio should be calculated over years, not months or days.

    Perhaps you already know this.

    There are plugin's though for excel that calculate a lot of these. I can't remember where I ran past them though. If I can find them again I will post in this thread.
  9. bone

    bone ET Sponsor

    Well, for Geeks. The rest of the universe - including the whales looking to place funds, relies upon the Sharpe.
  10. tyrian


    #10     Aug 15, 2011