What method would you recommend for calculating the daily stop-loss for a day trader? Some suggest using average PnL of winning days but I was wondering how you approach this aspect of the trading plan. Do you implement weekly/monthly stop-loss as well?
Depends how much your strategy makes. If plan A yields 20% a year, then a 2% daily stop loss would lead to ruin. If plan B yields 4% a day, then a 2% daily stop loss may actually be totally fine. After compounding.
Depends on whsat you anticipate for the current tradng day.You could do a % movement on the day or something if you wanted something more dynamic, just rebase line the opening price every day.
nothing more that 10% of your trading capital. I would have a 10% computer stop. and have a 7.5% mental stop to just slowly start covering my positions.
I would not use daily stops, but stops based on a number of trades. I calculate my risk based on a big number of trades. So my stats give my risk in function of a number of trades, not in reference to days. Some days you trade more than other days, so to keep the risk equal you should not watch days but number of trades.
The issue here is how to come up with a specific number. 2% per day? Why not 2.5%? I agree this has to be based on historical performance though.
"In exchange for universal applicability, you give up perfect execution" -Julia Ormond ( chartered technical analyst)
I agree. Daily limits tend to be psychological crutch used by amateurs. They don't tend to be derived using any logical method. A professional approach is to backtest your trading model and then use Monte Carlo analysis to derive criteria for determining if the model is operating within statistical expectation or not. If it has deviated from expectation by a statistically significant amount then common sense dictates you don't trade the model until you understand why it has deviated.
This is a good topic. Why not employe MAE, MFE to make the determination ? Compute your MAE and MFE daily for at least 100 days. Then focus on the average MAE as your daily stop loss figure and determine that if you quit at that figure, what percent of the days went on to be winners....and then vice versa for daily profit take figure. If the percentage is high, keep moving the figure lower (loss) and then higher (profit) towards 1 std deviation of the MAE, MFE. This method will also help you determine on which days you should quit early with a sizeable profit in hand.
I've never used a daily stop loss. Always a per trade stop loss. I know that no matter how many times I trade, I'll win more than half. Let's say 60% for this argument. My average win is "X" and my average stop loss is set to 0.5X. So it doesn't matter how much I trade, the math says I win. If I make 100 trades in a day (not likely since the most I've ever taken is 6), each one is independent from the last with it's own stop at roughly half of the expected return. The way I determine if I should stop trading for the rest of the day or keep going is an assessment of how level my head is. If I'm self destructing and making bad decisions I'm done. But if I'm just losing because of normal reasons, there's no reason to stop.