What a +0.84% Trailing-Stop Exit Taught Me About Rule-Based Automation
This is not a recommendation to buy, sell, or use any particular trading strategy. It is a reflection on why emotional regret can be dangerous when operating a rule-based automation system.
The trailing stop sold at +0.84%.
It was a gap-down open. The trailing stop is designed to sell automatically when price drops a set percentage from the peak. At 09:04, the system sold. If a person had been watching the screen, they would have said: “wait just a little longer.”
But that instinct is the problem. Waiting longer could just as easily have led to further decline and a loss. The current setup is not designed to maximize profit — it is designed to structurally prevent loss. Locking in +0.84% is not a failure.
The RS filter turned out to be stricter than expected. Out of 20 stocks in the universe, only 2 or 3 qualified as actual buy candidates. The rest were filtered out by one of three conditions: MA20 gap too wide, MA60 in a downtrend, or price below MA. Even on a BULL regime day, more than 80% of the universe gets eliminated. At first the question was “is this right?” — now it reads as a signal that the strategy is being cautious in the way it was designed to be.
The 20% cash-reserve logic also did its job. Even with ₩2.1 million in available cash, the system did not deploy it all. The structure prevents indiscriminate momentum chasing. The system makes the judgment — not the operator. That is the point of this automation.
Two bugs appeared and were caught the same day. An RS-disqualified stock was appearing in the buy-candidate log alongside eligible stocks, creating confusion — fixed by adding an [additional-buy-RS-exempt] tag to separate the categories. A web export function was reading the key “result” while the actual stored key was “regime” — renamed immediately.
Neither bug was critical, but catching and fixing both on the day they appeared matters. It means the system remains observable even while running live trades.
More data needs to accumulate before drawing conclusions. The current phase is: watch and wait.
What this exit confirmed is simpler than the numbers: when a system sells at +0.84% because that is what it was designed to do, and the operator’s first reaction is “what a waste” — that reaction is the real risk. A rule-based system that gets overridden on emotional grounds stops being rule-based.
→ Related operations log: 2026-04-29 Operations Log → Project history: stock-auto-trade project
※ This content is a personal experiment record, not investment advice.
All investment decisions and their consequences are the responsibility of the individual.