How can I tell if my system is ROBUST

Quote from Aafwintb:

Gooday everybody,

I would like to know if my system is ROBUST, but I do not know what specific performance measurements to look at. I use tradestation 2000i for my backtesting.

Any thoughts would be greately apprciated. Thank you.

"Cut your losses and let your profits run"

IMO, the best source for an answer to this question is Stridsman's book:

"TRADING SYSTEMS THAT WORK"

It is a very serious book that is sometimes hard to follow but well worth spending time with.

Don't get fooled by the title, the book is about the ways to find systems that long term work and not about the trading systems used as examples in the book.

DS
 
Quote from progers82:
if you were just using those two, then i would want my result to be much higher that 1 say in the 2+ range. I use all three and I use the 1 because when I am developing a trading system, it is so easy to be conned into believing something will work and most of the time it won't work into the future.

But what's the difference between your use of ratio of win/loss and % of win/loss? Seems like they're measuring the same thing - just different semantics.
 
PF and Avg Win/Avg Loss? [/QU

Is this a question or an answer/statement?

sorry about that. that was suppose to be my answer.

concerning another post:
they are not the same. if you have a positive profit factor and a less than good ave win/ave loss ratio it tells us that you could be profitable but you could experience loss because of a negitive event in the market with slippage. it means you are having to trade more to achieve that save profit. if you have a good PF and a good win/loss ration, it means you are not overtrading to achieve the same profit which reduces your risk because of your exposure to the market.
 
Quote from progers82:

PF and Avg Win/Avg Loss? [/QU

Is this a question or an answer/statement?

sorry about that. that was suppose to be my answer.

concerning another post:
they are not the same. if you have a positive profit factor and a less than good ave win/ave loss ratio it tells us that you could be profitable but you could experience loss because of a negitive event in the market with slippage. it means you are having to trade more to achieve that save profit. if you have a good PF and a good win/loss ration, it means you are not overtrading to achieve the same profit which reduces your risk because of your exposure to the market.

Thansk for the clarification/explanation. Why do you use 2 as a base line for the product of PF and Avg Win/Avg Loss? What is the range of values you've seen for this product, what would you consider to be a high value for a theoretically exceptional system?.
 
Quote from progers82:
[Bthey are not the same. if you have a positive profit factor and a less than good ave win/ave loss ratio... [/B]
That wasn't what I was asking.

Perhaps if you just give a clearer example of the three different values that you're using and their derivation it would clear up the whole question. With the brief text you used, it still looks like two of the three are essentially the same thing.

Thanks.
 
Quote from ron2368:

Post a system report.



How is your p/l distributed over your trades, if you have only a few trades that account for a majority of the profit that is bad.




RC

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

Think with my personality, Ron has an excellant point if its only one system only
which is clearly implied from the original question.

For those with & more than one system , in specificaly this or any bull market
not many trades /investments for majority of profit could be profitable & good with buy & sell plan.


Uptrends profits run ;
however just thought of a high % gainer uptrending semi
probably never will trade because
profits big trend in way too small & tricky a time frame,even over a year.

:cool:
 
Robust trading system should:

1. Have variables which produce reasonable profit that are not only your chosen variable. For example, you're using the 20 and 50 day moving average. What about the 19 and 49 day MA's? And the 21 and 51 day MA's? The system should also be profitable with them too. If there is a sharp drop-off in profitability with input parameters that are close to the numbers you are using, the system is likely over-optimized.

2. Have a reasonable drawdown, one you can live with.

3. Have some sort of stop so that you avoid having that 40% loss that every once in a while comes along and kills you. A 10% loss without much of a change in profitability will really make you feel better and keep you out of a free fall (provided you can get out, of course - remember, if in desparation you can always buy an option to hedge your position, at least temporarily).

Regards, S
 
Back
Top