This may surprise some people but stop losses usually have negative expectancy. Noticed this when playing with backtests and it's interesting to contrast with how they are described almost universally as positive. Yet, they are still useful. Why? Absolute advantage. 1/
Absolute advantage stems from a benefit one side gains from a trade. This makes a trade net positive and some of this gain can be shared with the other side, resulting in a win-win situation. For example, free trade benefits both countries. These also exist in financial markets.
One example: commercial hedgers locking in good prices to reduce business risk. This usually takes the form of selling into a rising market or buying into a falling one. The trader on the other side is paid for providing a service bearing that risk. Momentum in commodity markets.
Heard this described in this excellent @choffstein podcast https://blog.thinknewfound.com/podcast/s3e6-jeffrey-baird/
Also happens in stocks. A big winner can turn into an huge position and due to Kelly sizing, reducing risk is smart even if the stock remains a good buy. If selling is -EV then buying is +EV. Momentum in stocks.
Those who sold TSLA at $695 on the close for index inclusion also got paid for providing a service. It's gone a lot higher since, but sellers had a nice profit 1 minute later. Index funds needed to own it no matter what so the trade was a win-win despite having to overpay.
Back to stops. Large drawdowns will utterly demolish compound returns. Stops reduce the chance of a catastrophic loss, as well as preserving mental capital and guarding against biases like anchoring/loss aversion. But you are paying for these benefits.
I think this tweet puts it well. You are helping someone out and getting paid for it. It's a win-win because you make money and they don't blow their account. https://twitter.com/WallStCritic/status/1008768147568988160
Notice how in the first examples, you are paid to go with the trend, while with stops you are paid to go against it. I believe this shows up in the chart as when something goes "parabolic." Someone needs to buy it, urgently, possibly for self-preservation.
To sum up, look for absolute advantage trades, win-win situations where someone is receiving benefits other than money, if your own goal is to make money.
I should elaborate more on stops. If stops were +EV, you'd be able to build a system where you constantly get stopped out and make $. But how do you even do that? Stops can be set at any price. They only make sense in context of your trade and portfolio. They are personal.
A bad stop is one where a lot of other stops are, and you get a terrible price. A good stop will just result in random trading. 1987 was basically caused by a bunch of stop losses all on top of each other.