Every field acquires folklore, and retail algorithmic trading has more than most, because the product being sold — effortless money — is exactly what buyers most want to believe in. None of the myths below require malice; most spread through honest enthusiasm. All of them cost money.
The myths
- "Set and forget." Automation removes execution, not supervision. Strategies decay as market structure changes; someone has to notice. The realistic promise is minutes of oversight per day instead of hours at the screen.
- "More trades means more profit." Each trade pays costs. High-frequency retail strategies often generate impressive activity whose profits flow, in aggregate, to the broker's spread.
- "A 90% win rate means it works." Win rate without size is meaningless. Ninety small wins are erased by ten large losses. Expectancy — average result per trade including losses — is the only summary number worth quoting.
- "It uses AI, so it adapts." In this market, 'AI' usually labels ordinary rule-based code. Genuine adaptive systems exist and are harder, not easier, to validate — an adapting strategy can adapt itself into ruin. Treat the word as a prompt for harder questions, not comfort.
- "Backtest profit predicts live profit." Lesson four in one sentence: a backtest constrains what a strategy is, never what it will earn.
- "Recovery logic means it never loses." Martingale-style recovery converts many small wins and rare catastrophic losses into a curve that looks smooth right up until it isn't. Sometimes acceptable, never free — know when the rare day costs.
- "If it were profitable, nobody would sell it." A lazy dismissal in the other direction. Strategy development and capital are different resources; selling tools is a legitimate business, as it is in every industry. The correct posture is neither cynicism nor faith — it is verification, which this curriculum has now equipped you to do.
The honest summary
Automated trading is real, learnable, and unforgiving. The machinery is genuinely useful: it executes without fear, works while you sleep, and applies rules with a consistency no human can match. What it does not do is remove the requirement that the rules be worth executing — or the requirement that you, the operator, understand risk, testing, and costs well enough to supervise it.
If you have read the curriculum this far, you now know more about the practical operation of automated strategies than the large majority of people running them with real money. That is a low bar. It is also, genuinely, an edge.