Strategy directory

Each guide states the rules a setup uses — entry condition, the data it needs and the market conditions it assumes — so that it can be tested rather than believed. They are educational descriptions of well-known approaches, not recommendations, and they deliberately avoid position sizing or targets, because those depend on circumstances AlgoInsight knows nothing about.

A backtest shows how a fixed rule would have behaved on historical data. It cannot account for slippage on illiquid instruments, spread widening in fast markets, brokerage and taxes, or the fact that you may not follow the rule after several losses. Testing repeatedly until a set of parameters looks good — curve fitting — produces results that rarely survive on new data.

Drawdown is the fall from a previous equity peak, and it is usually the constraint that ends a strategy in practice. Win rate is meaningless without average win versus average loss. Look-ahead bias is accidentally using information that was not available at the time. Regime change is the reason a strategy that worked for years can stop working without any error in its rules.