Backtest average return is the mean per-trade result across all of a ticker’s past BUY signals — the wins and losses blended into one number, expressed as a percentage of entry price. It describes the size of the edge, not just how often it appeared.
Average return is a simple mean over a small sample, and a simple mean is easily hijacked by a single outlier. If a ticker has five past BUY signals and one caught a huge rally, that one trade can pull the average up on its own — making the signal look far more reliable than the other four suggest. (Example, illustrative: four flat trades and one +50% win average out to +10%, which reads as a solid edge but rests entirely on a single trade.) Always look at how few signals sit behind the number.
Average return is a plain average of individual trade outcomes, not a compounded growth rate you would earn by reinvesting. It does not tell you what a sequence of these trades would have done to an account, and it says nothing about the order of wins and losses or the drawdowns along the way. Read it as a per-trade expectation, not a portfolio return.
Average return is the expected-value term the Kelly position-sizing formula needs — paired with win rate, it says how much edge, if any, a bet carries and therefore how large the position should be. It is emphatically not a forecast of what the next trade will return. Because it is noisy and outlier-prone, the sizing that consumes it is confidence-scaled and capped, and it should be read next to the risk-reward ratio for the downside it does not show.
It is a simple mean over a small sample, so one unusually large win can pull the average up on its own and make a signal look more reliable than the rest of its history supports. Always check how few trades sit behind the number.
No. It is a plain average of individual trade outcomes, not a compounded growth rate. It does not tell you what a sequence of these trades would have done to an account or what drawdowns occurred along the way.
No. It is the expected-value input the Kelly sizing formula uses, not a forecast of the next trade. Because it is noisy, the sizing that consumes it is confidence-scaled and capped, and it should be read alongside the risk-reward ratio.
Browse all S&P 500 tickers to see this metric applied to individual companies.
Educational research only — not investment advice.