In a recent post, I introduced a new innovation in WFO which I called Weighted Anchored WFO (WAWFO). This new approach and innovation derived from the many problems that I recognized traditional WFO has and attempts to address many of the problems I recognized inherit to WFO (but cannot address all, unfortunately).
The Weighted Anchored WFO decays older performance by weighting newer performance more heavily. In this way, it shares some similarity to adaptive strategy techniques I have developed but requires less strategy-level plumbing. WAWFO has the benefit of allowing a strategy to be optimized over all the data while assigning greater relevance to the more recent data. Because the mathematical weighting function and ability to use all the data, it alleviates the need to discover the appropriate in-sample window size while equally addressing the problem of stale data inherit to traditional Anchored WFO.
Fortunately, this technique does not require that an entirely new WFO engine to be created. Below, I implement it in Multicharts by computing a weighted Net Profit measure. The actual performance measure could be changed out or made more advanced but it demonstrates the technique:
variables: ExpProfit(0), LastNetProfit(0), LastTradeProfit(0), BarCounter(0); //strategy logic here BarCounter = BarCounter+1; LastNetProfit = netprofit; if LastNetProfit <> LastNetProfit  then begin LastTradeProfit = LastNetProfit -LastNetProfit ; ExpProfit = ExpProfit + LastTradeProfit*BarCounter; end; SetCustomFitnessValue(ExpProfit );
It is possible to look for robustness and profit characteristics in the OOS by using Multichart’s Matrix level optimization for a given strategy. While hardly exhaustive, in my findings, the technique shows significant promise: producing better OOS results then many of the more common fitness functions.
Perhaps more importantly, the rationale behind the method makes sense to me: the rationale of considering all the data but giving greater weight to the most recent results.
The author is passionate about markets. He has developed top ranked futures strategies. His core focus is (1) applying machine learning and developing systematic strategies, and (2) solving the toughest problems of discretionary trading by applying quantitative tools, machine learning, and performance discipline. You can contact the author at email@example.com.
Please log in again. The login page will open in a new tab. After logging in you can close it and return to this page.