The topic is everything about stop losses. A stop loss is a kind of exit criteria. Typically, a stop loss is defined as a MAE price based stop. For systematic traders, the optimal stop can be determined during back testing. I have seen that stop losses near extremes tend to work better then “comfortably sized” stops. Typically, with mean reversion systems one will find that a larger stop loss works better then a small stop because more losing trades can recover to break even or a small profit. Momentum trades, on the other hand, are more likely to benefit from a smaller stop simply because when one buys or sells on momentum the pricing is often poor and one is anticipating something to happen.
Philosophically there are a few different perspectives on stops. One perspective is that stops should be placed where the traders know they are wrong. This makes sense provided that the market is more likely to continue when you are wrong. In most cases, this simply isn’t the case. This is why some traders eschew price based stops simply because it is very difficult to determine a price where the market is more likely to continue. Another view and one that I’ve started to ascribe too is that the stop loss has no predictive quality whatsoever. Instead, a stop loss can be viewed as a measure of “risk tolerance”. Within that risk tolerance, I might adjust it slightly if I think there might be a price where the market is more likely to reverse.
In general, it is difficult to trade with stop losses because they increase trading costs and tend to convert open trade drawdown into closed losses. A real problem with most stops is they “lose information” about the market. I noticed that all my predictions regarding the direction of the market had value even when I made conflicting predictions. In my predictive work, I always made my predictions falsifiable. But, another way to think about it is that any signal has a rapid fall off in value but one that doesn’t go strictly to zero.
Without stop losses though, one cannot leverage their trades/systems for outsize return. The amount that one can leverage a method is directly related to the win ratio, risk per trade, and max drawdown. Decreasing the risk per trade will tend to increase the max drawdown as open losses that might have recovered will be marked into the closed losses. In general, one desires a low risk per trade and a low max drawdown to produce outsize returns. This is very difficult to achieve in most efficient markets. Most of the actual verified systems that do work instead take a high risk per trade to reduce the drawdown. These sorts of systems can produce outsize returns: however, the tail risk is high because one will risk a significant portion of the account each trade. A problem to be aware of is that such systems become somewhat subject to the historical max risk on any given trade.
This topic would be remiss without a few additional points. As you reduce the risk per trade, your trading costs go up. So, without rock bottom costs then it becomes difficult to overcome costs. The second is that the price based stop loss is not the only type of stop. One can, of course, attempt other sorts of exits such as a time based exits or using exit criteria. Those methods can work better then price based stops but again incur more risk and introduce possibility for future high tail risk.
Curtis is passionate about markets. He has developed top ranked futures strategies. His core focus is (1) applying machine learning and developing systematic strategies, and (2) solving the toughest problems of discretionary trading by applying quantitative tools, machine learning, and performance discipline. You can contact him at email@example.com.
Please log in again. The login page will open in a new tab. After logging in you can close it and return to this page.