If youâre a system developer what do you do when it all goes wrong ? Take ETF timer as a perfect example; you have 3 years of fantastic performance through a massive market crash , bear and bulls market. You have developed 3 systems around this market timing principle, you trade your own account for a living and you have 200 subscribers and youâre pulling in 20k a month in revenue.
Then it all stops, your system goes from a nice 30 degree up equity curve to a downward 30 degree curve!
How do you protect yourself from something like this? Have multiple systems using different algos ? Donât give up your day job?
I think if you have an automated system (or largely automated system), you should have enough backtest history to know when your system will underperform. You might even know the conditions under which it would crash. Anyone with an automated system should be able to clearly state what those conditions are.
If similar conditions re-occur, you are left with a decision to make.
Some will “stick with the program” and just ride the underperformance period out. Not a bad choice, but we’ve all heard the phrase that the market can remain irrational far longer than you remain stay solvent.
I like having a backup plan in case the market does start to behave “irrationally”. The backup plan won’t violate the underlying principles of my strategy (in other words, I’m a mean-reversion trader, so I won’t suddenly try to be a trend-follower). Accurately knowing when to shift back and forth is more trouble than it’s worth. My backup plan will be another form of mean-reversion trading.
Now, sometimes I’ll be right in applying my backup plan. Sometimes I’ll be better off leaving things alone. But since my own money is at risk with every trade, I prefer to take a cautious approach. I don’t like losing money, and if I can do something to avoid it, I will.
