When to Stop a Trading System

[trend2009]

There should be some really bad drawdown(s) in your back test already. Otherwise it is probably curve fit or not long enough.

This maybe the case with your back test (hard for me tell given your limited metrics).

If so anything more than the biggest drawdown that you saw in the last 16 years would be a red flag.

his theoretical dd is small

 

[Gambitman]

I believe it was Charlie Wright who said in a trade station video 25 years ago something to the affect of “when it stops making money in the environment it should make money.” 2.5x drawdowns are table stakes.

 

[David's faith]

Maybe improve the filters for entries oder wait for the underlying market to behave as expected (not trading in uncertain times f.e.)

 

[Businessman]

his theoretical dd is small

Is it? Its not clear to me that it is.

Take system "A", is the 162K figure the profit per year or the total profit over 16 years?
If it is 162K per year, then I agree, the largest drawdown is very small.

However if it is total profit over 16 years. Then the average yearly profit for System "A" is only 10K a year and the biggest drawdown is 7K.
A 7K drawdown for a system that makes 10K a year is not a small drawdown.
If so then his back tests could already be including some large drawdowns, which is good.

 

[trend2009]

Is it? Its not clear to me that it is.

Take system "A", is the 162K figure the profit per year or the total profit over 16 years?
If it is 162K per year, then I agree, the largest drawdown is very small.

However if it is total profit over 16 years. Then the average yearly profit for System "A" is only 10K a year and the biggest drawdown is 7K.
A 7K drawdown for a system that makes 10K a year is not a small drawdown.
If so then his back tests could already be including some large drawdowns, which is good.

16.2k is for 16 years I guess. But the profit factor is 2, quite good.

 

[SunTrader]

Everyone knows that.

So if you realize that then why bring up govmint-think that only wealthy individuals should invest in hedge funds. Anyone can buy stonks and lose their money, why should hedgie investments be treated any different?

 

[Q.E.D.]

Hello All,

Here are some example 16 year back test systems performance metrics for :

View attachment 299426

Questions:

1. When do you recommend to stop the system due to system losing too much money? I am thinking somewhere between 2-2.5 times drawdown

Thank you

IMO, those systems never should have been traded in the real world: nothing in above results is anything likely (possible) to occur in reality. Results are due to data-fitting, IMO.

 

[zdreg]

Everyone knows that.

When you write throwaway responses you deserve to be attacked or better still be ignored.
quote from xenomorph. I'm used to being attacked.
https://www.elitetrader.com/et/thre...iced-when-scalping.371071/page-2#post-5715760

 

[Businessman]

IMO, those systems never should have been traded in the real world: nothing in above results is anything likely (possible) to occur in reality. Results are due to data-fitting, IMO.

Agree, most likely curve fit.

The almost 60% win rate is red flag.

 

[Scataphagos]

Agree, most likely curve fit.

The almost 60% win rate is red flag.

Agree. 60% is very much "high end". Size of wins should be significantly higher than size of losses... or something is seriously wrong.