I have been using a simple inflection system based on weighted moving averages to determine the long term direction of the market. So far, it has worked extremely well in entering and exiting long, market based positions. I have also found that it works extremely well intra-day, using constant volume bars. Unfortunately, when I try to apply this to individual equities with daily bars, it completely falls apart. I can't seem to identify a good moving average without optimizing for each equity -- which, even over a long period of time, I find to be too much of an exercise in curve fitting and not based on any fundamental truth of the market. I think I must add an additional parameter for recent volatility as well as covariance to market and sector movements (and perhaps their volatility as well?). An excellent example would be CAT recently: I was alerted that it was undervalued at around $65. However, volatility has been so high recently, moving averages barely budge. I never ended up investing and seem to have missed the opportunity -- the stock quite simply moved too fast due to recent market volatility. I have come up with a system that seems to be excellent at screening fundamentals to determine undervalued picks, but my timing is horrible -- and when fundamental data can be up to three months lagging, timing based on price action is a must! Thoughts? Thanks! C
If you test your system on tradeable index products like ETFs or stock index futures, the results may please you. As time passes and you test lots of systems on lots of tradeables, you might discover that your systems work best on (asset class 1), not so well on (asset class 2), and really poorly on (asset class 3). This is great if (asset class 1) is your favorite class, the one you feel most comfortable with, e.g., individual stocks. But it is terribly frightening if your favorite tradeables are (asset class 3), the ones your systems trade the worst. It is even MORE frightening if the asset class you loathe and fear the most, e.g., commodities, is the one that your systems trade the best (asset class 1). Now you get to choose between (A) throwing away all your systems, starting over; (B) deciding to trade your most profitable systems on asset class 1, the tradeables you loathe and fear. Sometimes it isn't easy.
I guess sometimes it takes an outside view to force you to acknowledge something you didn't want to have to admit ... Thanks C