anyone know the proper definition of Risk of Ruin?

Discussion in 'Strategy Development' started by yoyo2000, Mar 20, 2004.

  1. yoyo2000


    i found there is something wrong with the definition in the book 'Smarter Trading'

    at the end of the Chapter 4,Mr. Kaufman gives a formula to calculate the Risk of Ruin(=ROR),but i feel puzzle about it:
    the formula is :

    MaxRisk := 0.1;-- it should be the percent of account

    AvgWin := AvgWinPct / 100;
    AvgLoss := AvgLossPct / 100;
    Z := winpct / 100 * ( AvgWin) - ( 1 - winpct /100 ) * abs( AvgLoss);
    A := sqrt( winpct / 100 * AvgWin * AvgWin + ( 1 - winpct /100 ) * AvgLoss * AvgLoss);
    PP := 0.5 * ( 1 + ( Z / A));
    risk_of_ruin := power((( 1 - PP ) / PP ), MaxRisk / A);

    but: as a positive system,Z>0,
    and A must be positive,so PP > 0.5, so (1-PP)<0.5
    so ( 1 - PP ) / PP )<1,
    and if A is fixed, (MaxRisk / A ) increases when Max Risk increases.
    as ( 1 - PP ) / PP )<1 ,so when (MaxRisk / A ) increases ,
    risk of ruin becomes smaller,

    it means the more the capital i use in a positive system ,the less the risk of ruin , is it funny?
    what's wrong with it? or wrong with me?

    PS: the example in P68 is incorrect, the result of A should be 0.02 instead of 0.0605.
  2. I do not have Kaufman's "Smarter Trading", nor have I looked at your formulae. But you seem to have trouble understanding that risk of ruin decreases with an increase in trading capital. Please explain why?
  3. yoyo2000


    from the beginning,and theoretically, i thought Risk of Ruin should be bigger as the position size increases.
    that's the reason why i feel puzzle to that formula......

    and the formula is above,it's said from Ralph Vince's book:portfolio Management Formulas,or P. Griffin's The Theory of Blackjack.
  4. yoyo2000


    and i am sure there is a conflict between common sense and the formula,could u see what's wrong with the way i though?it is below the formula.
  5. SumJurk


    Another book you might check out is "Portfolio Management Formulas" by Ralph Vince. He has about a 17 page chapter on it.

    An interesting quote from the chapter..."Nothing in the realm of money management could be more useless to the trader" (than the risk of ruin calculations)

    Personally, I'd say just try to keep a 1-2 (or better) risk/reward ratio. You'll have a better chance of not being ruined...:)
  6. OK, I have found the formula you are referring to in another Kaufman book ("Trading Systems and Methods", p. 616) and have copied it into a spreadsheet (see attached gif). When I refer to trading capital, it is "Investment", or B3. Clearly, as B3 increases, risk of ruin (B12) decreases. Try it out for yourself - spend the 2 minutes required to construct the spreadsheet, and then instead of referring to terms such as "position size", formulate your question in terms of cell references. Otherwise, there will be no progress.

    As for the error on p.68, it won't be the first in a Kaufman book. I'm still waiting for a reply to an email I sent the publishers regarding a blooper I found in "Trading Systems and Methods" :
    • ror.gif
      File size:
      16.6 KB
  7. Measuring this mathematically seems to require certain assumptions regarding the probabilty of success. Consequently, it is my considered opinion (as opposed to my unconsidered opinions) that it is impossible to accurately predict risk of ruin in the market.

    Conclusion: Just wing it......with someone else's money!
  8. lol
  9. Just to add another view, attached is a copy of the way I learned risk of ruin.
    • ror.gif
      File size:
      18.4 KB
  10. yoyo2000


    hi£¬Allencook,ur algorithm doesn't include the size of earning or losing money,it only includes the percent of win%.but when we calculate the expectancy,we all know the size of earning or losing money is a necessary factor,so,i choose the formula that in my post,but........

    and,Mr Subliminal,u decrease the ROR by increasing the investment,but,as the investment increases,with the Avg Profit and loss fixed,
    the avgwinpct(=Avg Profit / investment) and
    the Avglosspct (=Avg Loss /investment) decreases,
    A should decreases,so PP increases , and (1- PP)/PP decrease , so risk of ruin decreases.

    and,i also want to know how the republisher responds to ur blooper......
    #10     Mar 21, 2004