Donchian breakouts on stocks

[China Diapers]

I have been experimenting with Donchian breakouts on EOD stocks, results for long trades seem to be OK with the below criteria:

Current days close is higher than the highs of the past 20 days
Current day closed up with a wide spread
Volume is 150% or greater than the average volume for the past 100 days
Stop set 20 ATR
Exit when price closes below MA (experimented with various length MA’s)

But when I try and reverse the above (excluding the volume) for shorts, the results are pretty horrible. I then tried a 40 days breakout in the hope that would ensure I am in a downtrend, that didn’t work. I tried a 10 day breakout because stocks fall faster than they rise, that didn’t work. I noticed that moves down are often choppier than moves up and my trail was being hit quite quickly, so I tried a longer MA, that didn’t work. So I guess it doesn’t work.

Maybe it’s the stocks I am selecting. Anybody have any thoughts on this?

 

[anley]

You're trading in a bull market which is why the longs are working better than shorts.

The reverse will happen when trading in a bear market.

 

[China Diapers]

Well yes this is true we are in a bull market, but I have backtested going back 10 years or more in some cases. Shorts didn't work so good in 2008 or any prior bear markets.

Thought maybe this is a known issue and somebody would be saying "well of course Donchians don't work on the short on stocks", maybe not then.

 

[anley]

There's a lot of money still traded via Donch. methods but the trick is to filter the signals, that therefore is your goal but it's a hell of a nut to crack.

 

[China Diapers]

hmmmm, interesting.

 

[China Diapers]

Prorealtime, far from ideal.