is there a proof that there's a threshold duration in bollinger bands

Discussion in 'Technical Analysis' started by markleeds, Aug 15, 2011.

  1. markleeds


    Hi Everyone. Thias is my first post so I apologize in advance if
    I do anything wrong.

    Here goes: When bollinger bands are used as a pure mean reversion strategy, the windowsize used determines how quickly the moving average reacts to the actual series. I don't currently do pairs trading using bollinger bands but I was studying them recently and my question is the following:

    Assuming one doesn't have any stop losses or bells and whistles
    in their Bollinger Bands method ( no other indicators also ),
    is it possible to show mathematically, that, in a mean reversion
    strategy, once the trade duration is greater than the window size,
    it is no longer possible to profit from that trade. In other words,
    after windowsize periods, one can only lose money by staying in that trade. How much of course is unknown.

    My intuition is that, if y_t stayed EXACTLY the same for the next window size points after trade entry, one would break even and exit after windowsize periods. Therefore, if you're still in the trade after windowsize periods, there must be a current loss and you can't recover from it. Does that sound correct ? Thanks.

    Also, any academic references are appreciated also.