Friday, March 09, 2007

A New Guy In Town

Larry Connors, the guy who brought TradingMarkets to the web just opened up his own Blog and this is going to be good. This guy has probably forgotten more about trading than I'll ever know. I put a link over on the link list and you really should go visit frequently.

Today's inaugural column is about his and my favorite subject - regression to the mean - which is pretty much my bread and butter and yours too I'd wager. Anyway he has some statistical evidence that shows clearly that buying stocks off their 10-day low is a lot more profitable than buying them off their 10-day high. Well we know that of course. He also said that it would be better if they were above their 200-day moving average.

Sensing a challenge to my never ending quest to turn everyone against the 200 day MA I put together some filters - one using the 200 MA and one using the 90 EMA. Now these are not equivalent but the 90 is my favorite so I wanted to see how it would do. Not as good as the 200 MA. Well!

So I tried one other thing - I required the stock to close at its 10-day low and below the lower donchian band. The 200 MA still won and was hands down the best.

Here's that filter-
show stocks where close is between 15 and 35
and average volume(90) > 500000
and close reached a new 10 day low
and close > ma(200)
and close < lower donchian band(20,4)

And the results were win rate 72% (one trade a day) and ROI 78%. The ROI could be better but the net change over 30 days was a very nice 3.14%. So short term it is pretty efficient and long term it is pretty profitable.


Bullish Jim said...

Great results. Ironically, MSFT comes up today if you run that screen.


New faces same results.