Drawdowns – Have you got the Stomach??
- This topic has 19 replies, 8 voices, and was last updated 6 years ago by
.
-
Drawdowns can be very tricky to handle. Or better put, when to shut down a system in a DD period and when to let it run hoping for recovery, is hard to decide.
In my first half year of automatic trading I had read so often that one mustn’t turn off strategies after some losing trades, you have to give them a chance. Alright I thought and acted up to it (I thought) with the result that I almost never turned off a system even after several weeks of DD. This didn’t get me anywhere. I had profitable systems but the losers where always eating up all profits and after all I was losing. Not much but still.
At the end of 2017 I said to myself, that this can’t be the way to go and tried a new approach. I turned off all my losing systems and just kept the ones I had a good feeling about. This was a quite big step, I reduced the number of active strategies by about 40%. But so far I did not regret it. The strategies I kept performed very well even during the turmoil of the last months and I’m up for the year by more than 100%.
So the lesson is, yes you have to give strategies a real chance but at some point you have to turn them off. 🙂 But I’m still pretty unsure when this point is reached. For the time being I’m working with a test period of 6 months. If after 6 month of live trading the result does not approximately match what one would expect from the backtest performance, I turn it off.
This seems to work quite OK for strategies running on daily bars. Strategies that run intraday and trade much more often maybe need shorter time for evaluation but I’m not sure about it.
1 user thanked author for this post.
Here is an example of how having a simple filter that stops a strategy trading if price is below whatever price was ‘x’ periods ago can massively reduce drawdowns.
In this example the lookback period is 1400 hours or approx a quarter of a year. The orange line is buy and hold and it can be seen that there are some massive potential drawdowns if you had got into a trade at the wrong time and held on to it or carried on trading through massive market corrections. The strategy itself only does slightly better than buy and hold but it has long periods not in the market (zero risk time) and the drawdown is only around 6% with a starting balance of 10K. The filter does not close any trades, it just stops new ones being opened until price has stabilised or recovered.
It’s not exciting but it is much lower risk trading.
Here is the same strategy but with the lookback filter removed. Yes it is a little bit more profitable but the drawdown has increased to almost 10%. Although not a massive increase (maybe this was not a good example strategy to choose) it is an increase in drawdown. I guess it confirms that if you have the balance and the balls to sit through larger drawdown then you can make more money.
Hi GraHal, I cut my systems in February and have been looking for a solution to the uptick in volatility before restarting my algorithms. My systems are basic buy on support but don’t work when the market turns down very suddenly.
The solution I’m testing is a moving average ATR. I have a fast and a slow moving average and if the faster is above the slower then it doesn’t trade. It looks good on back tests and filters out the worst of the drops in February and March. I have it on demo and will let you know how things go.
1 user thanked author for this post.
Find exclusive trading pro-tools on 
