All of these points are irrelevant. If you have the data that shows 99.5% win rate, you would have the data on your expected win size, and could compute a fixed loss stop that would keep you on-side. In the example, which I'm assuming must be hyperbolic, your average win would only need to stay >1/199th of your average loss. I would agree that if a person cannot manage their positions to this, that person should stick to investing.

> your average win would only need to stay >1/199th of your average loss

You're not wrong, we're just talking about different things. You're talking about a hypothetical system that has a 99.5% win rate forever, and I'm talking about the way the market actually works.

In my version the market is often random, or to say it more accurately sometimes the Efficient Market Hypothesis is true and sometimes it's a Random Walk. His strategy likely worked during periods of stability when EMH held true. His strategy would have been 99.5% effective, until it suddenly started being no better than a random bet.

Of course he will have sized his stop loss for the times when his strategy was working, meaning he'll have placed it at something ludicrous like a 90% loss to maximize profit and avoid being taken out (anything else would have left profit on the table). This huge drawdown will suddenly wipe out almost all of his profits when the market shifts to random walk mode.

Even if he did size bets more reasonably, the length of time the market remains in random walk mode is itself random. Over a long enough time frame he'll eventually encounter random periods long enough to break him, even with more conservative sizing.

The market is adaptive, particularly with options, and will adapt to bust your approach in a heartbeat. It will in general adapt to bust a lot more sophisticated approaches. This is not physics. The underlying assumptions change.

In general, yes - no edge is forever. But that doesn't mean trading is a coin flip where no edge can exist. The market is adaptive because humans are adaptive - but, for example, their machine offspring are often slightly less adaptive and create opportunities for persistent returns. Even more so for most retail where your position size isn't creating slippage.

I can try a dynamic intermediate stop-loss point, with the intended goal that is not too early where it triggers often, and not too late where the loss mounts geometrically. This is intended to greatly limit the max loss per trade, which is good, but it still has significant risk of being triggered unnecessarily. I am highly skeptical that this will work, because I even see in the chart that it doesn't, but it's worth a test.

Yes, optimizations of this form can work to manage tail risk on an edge and can reproducibly lift expectancy for a strategy. GP is essentially claiming no positive edge can exist, which is false - though these edges may be short-lived and dependent on particular market regimes.