Strategy performances decline over time?

[TZT]

I have traded historical-data-backtested strategies since 1999, in stocks, futures, and funds. Once I have discovered and implemented a profitable trading strategy, its performances always got worse and worse over time. None of my strategies remained profitable after 10 years.

I am curious whether other strategy traders have seen the same pattern? And how your guys manage this issue?

For me, I generally start to reduce my position in a strategy if it is unprofitable for a year. If the strategy remains unprofitable for another year or two, I stop trading it altogether. [My strategies usually generate dozens of trades a year; so a couple of years give me enough sample to suspect that the world has changed.]

 

[fan27]

I have mostly automated finding, optimizing and combining strategies into a strategy portfolio. What I do is once a quarter I run the entire process for finding, optimizing and combining strategies. As part of that process, I have filters that discard strategies that do not meet certain performance metrics and I pick the best ending strategy portfolio. The result of this process is that strategies that are no longer performing get automatically discarded and new strategies emerge. So far (6 months of live trading) the process is working out.

 

[Scataphagos]

I have traded historical-data-backtested strategies since 1999, in stocks, futures, and funds. Once I have discovered and implemented a profitable trading strategy, its performances always got worse and worse over time. None of my strategies remained profitable after 10 years.

I am curious whether other strategy traders have seen the same pattern? And how your guys manage this issue?

For me, I generally start to reduce my position in a strategy if it is unprofitable for a year. If the strategy remains unprofitable for another year or two, I stop trading it altogether. [My strategies usually generate dozens of trades a year; so a couple of years give me enough sample to suspect that the world has changed.]

You have observed a great lesson about the markets..... "they change over time".

For you to be a long-time success, you need to adapt and change as the markets do.

Not so easy. Even big-time names most often fail at this.

I've always had the notion... "All you have to do for the market to knock you on your ass, is to keep doing that which was successful in the past".

Is there a solution to all of this?

Yes. It's learning "Price TA*". It's the one discipline which learned and properly applied CANNOT BE FAKED-OUT BIG TIME!

*Much to be learned and not so easy to apply well.

 

[tommcginnis]

If the markets are efficient (in that classical sense), then your job is to find the passing inefficiencies and *exploit* them via arbitrage: your *job* is to help the market run better.

If the inefficiencies that you find were perpetual, then everyone would (somehow) have similarly found+exploited them -- 'solving' the same problem again and again: a situation that is logically impossible. The problems you find -- whether an underpriced stock, or an overpriced index, or a too-widely priced options spread -- are by construction and by definition, impermanent to the point of being ephemeral.

If you don't do as fan27 does, and test, retest, cycle through and feedback around.... you'll be picking fruit from a long-dead tree. "Ewwww!"

 

[carrer]

1. Strategy based on market behavior. Certain markets have certain behaviors, it's own behaviors. These behaviors will probably not last long, or it could be long. It depends on how long the herd keeps trading the same way.

2. Strategy based on randomness behavior or random walk. This will work on all markets and indefinitely. I have my strategies running in this category as it works on various markets and it could be diversified easily. Although the return is lower than the 'number 1' strategy but I am happy with it, without having to worry when it will stop working, hence, not having to find a new working strategy.

 

[Scataphagos]

1. Correct about your #1, except it's not "certain markets", but rather "all markets".

2. How do you base a strategy on "randomness"?

 

[carrer]

1. Correct about your #1, except it's not "certain markets", but rather "all markets".

2. How do you base a strategy on "randomness"?

1. Say a strategy that works on S&P500 index will probably work on Dow but will not work on EURUSD.

2. Basically it's trend following.

 

[d08]

You have observed a great lesson about the markets..... "they change over time".

For you to be a long-time success, you need to adapt and change as the markets do.

Not so easy. Even big-time names most often fail at this.

I've always had the notion... "All you have to do for the market to knock you on your ass, is to keep doing that which was successful in the past".

Is there a solution to all of this?

Yes. It's learning "Price TA*". It's the one discipline which learned and properly applied CANNOT BE FAKED-OUT BIG TIME!

*Much to be learned and not so easy to apply well.

I agree with everything up to the Price TA part. Everything changes, even that. It's most likely just because the changes are subtle and appear over a long period, you don't notice them.

 

[qlai]

As part of that process, I have filters that discard strategies that do not meet certain performance metrics and I pick the best ending strategy portfolio. The result of this process is that strategies that are no longer performing get automatically discarded and new strategies emerge.

I was listening to a podcast recently and the guy described similar logic. One of the things I found interesting that for some strategies, he actually would only start trading them live when they went into draw down on sim. My understanding he is trading all of them all the time in the sim but chooses which to start or stop trading live based on market conditions.

 

[d08]

I was listening to a podcast recently and the guy described similar logic. One of the things I found interesting that for some strategies, he actually would only start trading them live when they went into draw down on sim. My understanding he is trading all of them all the time in the sim but chooses which to start or stop trading live based on market conditions.

That requires a lot of confidence in the strategies. I've seen them go into a drawdown, then into deeper drawdown, then record drawdown and then sideways for years.