Every Strategy Has Its Day

[Here2learn]

Some good advice here, thanks.

Most of my strategies are intraday, so they find signals every day. The problem is that they run a week or month with relatively consistent gains, then go stupid and loose for a week or month at a time.

Granted, some of this is poor strategy design. But until I can learn/find a teacher to help me develop better strategies, I can only work within my own abilities. So my goal for now is to identify the periods when the strategies work and trade them, then sit on the sidelines when I suspect they are going to be unsuccessful.

Looking at volatility is sage advice... any suggestions in the commodity markets? I trade mostly grains. I'll be thinking about the True Range concept and how to incorporate that into my model.

Thanks again.

 

[oldtime]

Quote from Here2learn:

Some good advice here, thanks.

Most of my strategies are intraday, so they find signals every day. The problem is that they run a week or month with relatively consistent gains, then go stupid and loose for a week or month at a time.

Granted, some of this is poor strategy design. But until I can learn/find a teacher to help me develop better strategies, I can only work within my own abilities. So my goal for now is to identify the periods when the strategies work and trade them, then sit on the sidelines when I suspect they are going to be unsuccessful.

Looking at volatility is sage advice... any suggestions in the commodity markets? I trade mostly grains. I'll be thinking about the True Range concept and how to incorporate that into my model.

Thanks again.

well, I dont usually give free advice, since most of it is not worth a damn anyway, but here's a cheap way to do it.

Run the same system in your paper account, and when it is really performing badly put it on in your real account.

So basically, you are just fading yourself.

Like the man said, "No one ever went broke underestimating the stupidity of his fellow man." (which in this case is you.)

 

[ocean5]

Quote from Here2learn:



..then go stupid and loose for a week or month at a time.

It may mean that you are wrong with the direction and keep adding to the loser every tick against you.Disaster.That happens,when you are wrong with the direction and you realize that it will go further against you.What can be done..Either close it right there,or if you want to add,there is no fast thinking there.Add when in turns in your direction.
For e.g. if i went wrong with such instrument like CL,i`d add not less then every 0.40,then 0.60,0.80 cent to get out with a scratch-1 tick plus.

 

[savagemp5]

Quote from dom993:

I would say if your strategy is about BE in "bad" (for that strategy) market conditions, and makes decent money in "good" conditions, then your best bet is to trade it regardless of market conditions.

But ... if you can forecast bad market conditions with quasi certainty, then it is better to turn your strategy off during these times.

Example1: most systems perform poorly in low-volume conditions, and you can forecast the week in-between Xmas & new-year is surely a low-volume period (just like the last week of august, up-to labor-day).

Example2: most systems perform poorly on options-expiration day.

Example3: you might find that your system doesn't do well around news events - analyze your backtest results for entries in 8:30am-8:45am & 10am-10:15am windows.

Example4: you might find your system doesn't do well the morning - or the afternoon - or during lunch-time (another low-volume window).

I would say this is probably one of the best advice, as im doing this myself. Of course there's alot more to "filter" the good and the bad.

Problem is when you design a strategy and you do not know the conditions as to why it makes money sometimes and doesn't other times, you are shooting yourself in the head.

My Examples:
#1 - I dont run my systems everyday Friday as I realize it loses more than other days.

#2 - Flat futures day. I dont run my systems when vix or futures are not higher or lower on that day. U figure out how to integrate this.

#3 - No movements for individual stocks. Tons of TA to help you ID the ups and downs.

#4 - Time of day, as mentioned. If you have multiple systems you will, if attentive enough, (you are suppose to know what you are doing!) know which strategy works well given the start/end time of day.

Heh tons of advice given for so many years of hardwork

 

[indicator777]

Quote from Here2learn:

In my experience, there are many good strategies around, but they are only good under the right market conditions.

Your experience is correct.

 

[oldtime]

Quote from indicator777:

Your experience is correct.

yes, but "right market conditions" are part of a good strategy

so the strategy is good 24/7 365 days a year, but it takes into account market condidtions, and either stays flat or adjusts

 

[dom993]

Quote from savagemp5:

I would say this is probably one of the best advice, as im doing this myself.

Thanks ... since those kind words were for me, I feel allowed to warn you about what's know in psychology as "confirmation bias"

http://en.wikipedia.org/wiki/Confirmation_bias


To the OP, the best use you can make of all advices is to take them as inputs in your research process, and get to your own conclusions based on your own results.