Inside bar breakout trades 102 times per ticker and ends up flat

Inside bar breakout generated 21,868 completed trades across the 214 tickers that finished a run, and its average return per ticker is 0.03 percent. That is not a rounding artifact of a quiet strategy. It is the most active strategy on the whole site, by a wide margin, and it still ended up almost exactly where it started.

21,868 trades over 214 tickers works out to roughly 102 trades per ticker across the ten year window, about one trade every five weeks, on every one of 214 stocks, for a decade. Its aggregate risk to reward ratio, gross profit divided by gross loss across every trade, is 1.0005, close enough to the 1.0 breakeven line that a difference of one basis point either way could flip which side of it the strategy sits on.

There is no line to draw through that chart. NR7 volatility breakout trades about 2.2 times per ticker and returned negative 0.09 percent. Golden cross trades about 6.8 times per ticker and returned 9.02 percent. Williams %R reversal trades about 39.7 times per ticker, five times as often as golden cross, and returned 27.21 percent, the best of the six. Inside bar breakout trades more than twice as often as Williams %R reversal and returned 0.03 percent. A reader hoping trade count would explain any of these results, in either direction, would be reading a number that does not carry that information.

A large sample saying nothing is still worth reading correctly

Tenachine's guide on a low win rate makes the case that win rate alone tells you how often a strategy wins, and risk to reward tells you how much, and a rule needs both read together before it means anything. Inside bar breakout's risk to reward of 1.0005 is the clearest possible illustration of that second number doing its job: the strategy's wins and losses are, in aggregate, worth almost exactly the same amount, so no win rate, however high or low, could have rescued the result. A 47.3 percent win rate elsewhere on this site, stochastic oversold bounce's, means something different than a 36.36 percent win rate here, precisely because the payoff behind each one is different, and inside bar breakout's payoff is close to a coin flip.

There is one thing 21,868 trades does buy, and it runs the opposite direction from the small-sample warnings elsewhere on this site. Tenachine's guide on the Kelly criterion treats a 14-trade record as too thin a base to trust a formula's output. A result built on 21,868 trades across 214 different stocks is a different kind of number: whatever inside bar breakout's true long run edge is, this sample is large enough that 0.03 percent is a stable estimate of it, not noise that a few more trades would likely overturn. The honest reading is not that the sample is too small to trust. It is that the result is trustworthy and the result is nothing.

A strategy trading roughly ten times a year on every ticker it touches would also be unusually exposed to whatever a live account's execution costs actually are, and every study on this site carries the same disclaimer: real world execution costs, slippage, and survivorship bias can materially change outcomes. A backtest already sitting at 0.03 percent before those costs are subtracted is not a strategy that a small improvement in execution turns into an edge. It is closer to a coin flip that traded itself to a standstill, 102 times per ticker, for ten years.