Performance decay on equities strategy

[TraDaToR]

Hello,

In the last 6 month I traded my first profitable stock strategy. It's half automated, trades a portfolio of stocks and trade duration is between 5 minutes and 1 day. At first I was really pleased with the consistency of returns. I was fine tuning it, looking for better instruments, slowing scaling up and thought : "I have something worth pursuing here"

Then around Christmas( around day 85 ), I thought I had worked enough on it, started to look for other strategies and focus back a little on futures( which is still my bread and butter... ). I stopped refining it and selecting stocks on a weekly basis. Here is the result :



The decay is impressive and it was easy to feel that other market operators adjusted quite fast to my orders.I have been forced out of markets a few times before on futures, a few times I opened my platform, placed my orders and immediately thought "OK, here it is. This algo simply put my strategy out of business". But I had never seen such a slow consistent decay on a whole portfolio of instruments.

Thoughts?

 

[stevegee58]

6 months of results isn't a good sample to judge. Maybe your strategy was simply profitable due to market conditions and then conditions changed. Lots of people felt like geniuses during the 90's.

 

[Scataphagos]

"Things" work, then they don't. Maybe they come back and work again, maybe not.
Maybe you need a different thing when the old one stops working.

It's the way of the markets.

Adapt or perish.

 

[wintergasp]

Hello,

In the last 6 month I traded my first profitable stock strategy. It's half automated, trades a portfolio of stocks and trade duration is between 5 minutes and 1 day. At first I was really pleased with the consistency of returns. I was fine tuning it, looking for better instruments, slowing scaling up and thought : "I have something worth pursuing here"

Then around Christmas( around day 85 ), I thought I had worked enough on it, started to look for other strategies and focus back a little on futures( which is still my bread and butter... ). I stopped refining it and selecting stocks on a weekly basis. Here is the result :



The decay is impressive and it was easy to feel that other market operators adjusted quite fast to my orders.I have been forced out of markets a few times before on futures, a few times I opened my platform, placed my orders and immediately thought "OK, here it is. This algo simply put my strategy out of business". But I had never seen such a slow consistent decay on a whole portfolio of instruments.

Thoughts?

So you've been flat for 20 days and you think it doesn't work? Just look at any other hedge fund, people have drawdowns for 1 year, 2 years, and are still profitable in the long run.

What you should look at is how robust your approach is, what was your research process, what was the room for too much fitting, what's your expected return per volatility etc...

If you look at your chart, between day 1 and 20, you've been as flat as between day 115 and 130

 

[TraDaToR]

I still think it is a profitable strategy. I am just impressed at how fast other participants have adjusted to my presence. Faster than on futures I think. On some of the most profitable instruments, I now miss fills by 1 tick which means one other participant just improves my ask or bid by one tick. Between day one and 20, I was just trading 100 shares on a limited number of stocks.

 

[TraDaToR]

6 months of results isn't a good sample to judge. Maybe your strategy was simply profitable due to market conditions and then conditions changed. Lots of people felt like geniuses during the 90's.

6 months but easily something like 500 trades. Each point is a daily P&L. I don't mean it is really significant but there was definitely an edge there when I was searching for suitable instruments and monitoring the portfolio actively.

 

[R1234]

Could possibly be a function of overall market dynamics and not specific to your strategy?
A broker I was recently speaking with said that several CTAs doing intraday strategies stopped working during Q1.

 

[murray t turtle]

6 months but easily something like 500 trades. Each point is a daily P&L. I don't mean it is really significant but there was definitely an edge there when I was searching for suitable instruments and monitoring the portfolio actively.

%% TRAda;
Even if that is almost 200 days, which is a good number.Sorry i have to agree with Steve 58.And if you want a merry Christmas, according to your chart i would NOT increase size, on 6 months. I , could design a system with profits on 2 days average trade; but 58 days goal, tends to be much better profits, with much better [lower]costs.

And if i was wanting to make money on the short side; unless its was run away bear market trend i may not trade Christmas/+. Even a turkey like TGT may fly up .