Hey guys , currently back testing strategies and my win% are high , but I really want to get my max drawn downs lower but I cant seem to get any lower than 45%? Anyone been stuck at this also? The only indicators I can test at the moment are MA/EMA/BB
I would not worry about those draw downs. Just wait till you test back 2 years on tick by tick data looking inside the bars. Your indicators will indicate something totally new to you!
The best way to limit drawdown is by getting in and exiting fast with small profit objective and even smaller stop. But in this case MAs won't help you. You need to move to something with less lag. You sound to me like me 10 years ago. You'll figure this out alone in time.
1) Do you attend NIU or U-Dub? 2) The fact that you're using "indicators " will be a big "problem". 3) Traders tend to optimize an indicator(s) to recent market activity. When a some type of change occurs in the activity, "large" drawdowns can result if you're too slow to exit losers or change the indicator(s). 4) The fact that you have a "high" percentage of profitable trades is potentially dangerous because you might be losing too much money on losing trades, that occur a "small" percentage of the time, that contribute to the large drawdowns. :eek: 5) If you're going to continue using indicators, you have to hope for too much luck and smoother trend-breakouts to keep you profitable. 6) To trade a shorter time-frame can be beneficial in getting you out of losers sooner. Otherwise, you'll always be chasing after the optimal setting for the indicators.
drawdowns % of what? I would suggest using only $-values when you assess a strategy, as this is what's in your account, not % The most important figure when you evaluate a strategy is the avg net per trade (using a fixed position size). Of course, with stocks, you want to make that rather a fixed position value, and derive the actual position size from that amount, divided by the stock price at the time of entry. If you want to normalize drawdowns, then do it in relation to the average net / trade. For example, if your average net per trade is $150, and your max drawdown is -$15,000, then this max drawdown is 100 avg/net ... and it becomes easier to compare drawdown between strategies. Another important aspect of drawdowns is the time to recover ... rather than "time", look at the number of trades to recover (from drawdown, to a new P&L peak). Anyway, reducing drawdowns requires improving the expectancy of your system (in other words, its average / trade). You can do that by either increasing your win%, or increasing the avg-Win/ avg-Loss, or both. I'll give you a short-cut ... instead of backtesting / optimizing trading systems, spend time doing statistical analysis, and finding the best expectancy "patterns".
Agreed. The fact that your win rate is high but your Max DD is also high is concerning. These types of systems have a tendency to blow up. The fact that you can only test those three indicators is most likely not helping you. Everybody else looks at those indicators as well...that's problematic. Your testing software probably included those preprogrammed for free, you will get what you pay for in this case.
i'm a discretionary trader. i had my big breakthrough when i minimized my largest losses, so i know what you're going through. i found a few different things that happened when i had my large losses but the ones which were most important are: 1) i ignored my rules and either went in too soon / put on too much size / tried trading differently than i normally do OR: 2) (and most appropriate with your situation) found that the big loss trades had identical setups to my normal gain / big gain trades....BUT, they all started acting differently than they should. that is, if the stock 'should have' started going up/down within the first 60 minutes, and didn't...most of the time it would be a bad trade. so instead of waiting until the stock took out the highs/lows, and everyone else was piling in and i'd get steamrolled...or instead of adding to a losing position - i would take a chunk OFF if it wasn't working when it should have. if the stock then became a good trade, i'd still have some, or could put more on once it started acting the way i think it should. but if it was a bad trade...well, i've now cut back my size considerably and can more easily exit the trade without getting squeezed or trampled.
Not that I do it, but I backtested a while ago. on asset classes, only be long in when price is > 10month MA. Reduces drawdowns in half. Tests out in high yield and equities quite well.
Definitely test this... You'll probably discover that winning trades are profitable from the start. Create and test a bunch of filters in the first 5-20% of your average trade length, and see what your results are if you exit the trades showing immediate losses.. Also put in some logic so you'll reenter the trade once it starts acting right again. Sometimes, you'll find you're too early... No sense missing the big move of a trade that "eventually" worked. Your win-rate will go down, but your expectancy should go up. And your DD % should drop considerably. Assuming you don't get hit by a black swan... If your expectancy and DD do not improve, then maybe the strategy sucks, and you should scrap it for something else. Advance only when you're sure you won't have to retreat.