Expectancy above .50??

Discussion in 'Risk Management' started by shanoballs, May 7, 2006.

  1. man


    totally agreed. use sharpe and you got what you want. use sortino if you want to be more aggressive. you can compare systems directly, no matter what their frequency is.

    as for the tradestation graphics someone postes bevore. again a source of trouble. it does not show times of much are little activity. it is again not what you will finally "feel" ... the charts are misleading as is most TS output. nevertheless experienced people overcome this by learning how to read the TS output by combining different bits and pieces.
    #21     May 16, 2006
  2. Let it be known that in the past some very smart people at ET have argued exactly the opposite: believing the Profit Ratio to be preferable to the Sharpe Ratio. (you could do a search)

    As for myself, I find a quick look at my Profit Ratio useful, at my Sharpe Ratio less useful. I use in fact a much more sophisticated criterion that took me years to develop.

    #22     May 16, 2006
    #23     May 16, 2006
  4. man


    did a (very quick) search, did not find confirmation of your statement. i personally found people on ET that did very well, yet ignored sharpe completely. but i found none who was familiar with the concept yet perceived it as of less value than profit factor.

    actually i think one of the biggest flaws people suffer from is wrong utility function. i think it is great that you took time to develop it! though i admit i still do not honestly consider to change mine. reason for that is that i want to be match my personal utility function. i hate too much volatility. naturally i do not like it downside but i get concerned with too much upside as well. at the moment we do print much in trend following, but i prefer stable 2%- to 10%-months ... but that is just me ...
    #24     May 16, 2006
  5. man,

    No arguing about this. You should use what works best for you.


    #25     May 16, 2006
  6. man


    traders think in dollars and pnl, fund managers think in % and equity curve figures. it is more about professional OPM managers than investor's point of view. 90% of investors do not know what to do with all these figures anyways ...
    #26     May 16, 2006
  7. Totally disagree. If I were limited to one metric I would choose profit factor. It's elegant in its simplicity and tells you how much you can expect to gain for every dollar lost. For example, a PF of 1.84 means you can expect, on average, to gain $1.84 for every $1 lost. PF = Total Profit/Total loss. Another way of looking at is: (PW * AW) / (PL * AL), where PW = Probability of a win, AW = Average win, PL = Probability of a loss, and AL = Average loss.

    There are very few systems with a PF of 1.25 that I'd prefer over one with 2.5. Things like the system with 2.5 making all of its gains from a single trade. But I can give you examples like this for ANY single metric. So of course it must be looked at in conjunction with other metrics.

    Profit Factor is not an entire equity curve in a single number, but neither is Sharpe which really is flawed BTW, because it penalizes large positive returns.
    #27     May 23, 2006
  8. man


    well, sharpe tells much more about an equity curve than does profit factor. and if you don't want to penalize upward vola then
    use sortino. nevertheless, profit factor IS flawed.
    #28     May 24, 2006
  9. Personally my sharpe ratio is 5.2 at the moment....and I am typically penalized by up vol... So it seems to be possible to reach something higher than 1.
    #29     May 24, 2006
  10. man


    if it is for longer than 2 years and we are not talking short any kind of option, then i'd say that 5.2 is a level where sortino and sharpe won't differ that much. high sharpes usually show low deviation in either direction.
    #30     May 24, 2006