Question for the Stats Gurus

Discussion in 'Professional Trading' started by TGregg, Jan 25, 2004.

  1. TGregg


    Doing my year end analysis, and I think I may have found something meaningless. :D

    My profitable trades average about 17 minutes in length with a standard dev of 26.5. That means there is not a significant meaning to the trade length, right?
  2. TGregg


    BTW, Correlation comes out to -.31.
  3. Aaron


    From the stdev being greater than the mean, we can deduce that you had mostly pretty short trades (a few minutes) and a few rather long trades (an hour or more).

    Unless you are doing a hypothesis test of some sort, there is no "p-value" that you are testing the significance of. So, I'd say you are right — in the narrow sense of statistical significance, the average and standard deviation aren't particularly meaningful.

    You could, for example, compare the average trade length for profitable versus unprofitable trades and do a hypothsis test and calculate a test statistic and measure its significance if you like. That's how statisticians use the word "significance".
  4. Aaron


    The correlation between profit and minutes held? If that is true, then that is interesting. It means, in general, the longer you hold a trade the less profitable it is. You should look for a cause and effect relationship. Are the longer trades less profitable because you held them so long? Or maybe it was just that you held them so long because they weren't so profitable. Do you trade with a profit target?
  5. TGregg


    Thanks for your help, Aaron.

    Yes, that is correct. There is a small (?) but negative correlation betweem minutes in a trade and profitablity.

    Yes, I trade with a profit target.