Trend-beta

A long-only strategy can make money in a bull market almost no matter what its rules say. The market did the work; the strategy just rode along. Trend-beta is what's left of a strategy's returns once you subtract what a passive position in the same instrument would have earned over the same window. Strip it out and a lot of "systems" turn out to be buy-and-hold wearing a costume.

The mechanism explains itself once you see it. Most trend-following, breakout, and momentum rules are structurally long-biased: they buy strength and hold through pullbacks. Backtest that construction over a multi-year uptrend and the equity curve climbs beautifully. The rules aren't predicting anything. They're just riding drift. Buy-and-hold would have matched or beaten the result, with none of the entry logic and none of the fees.

The test we run is deliberately unglamorous. Take the strategy's equity curve and compare it against simply holding the instrument flat over the identical dates. If the two tracks move together, same shape, same drawdowns, same up-years, the strategy added nothing a single buy order wouldn't have. Trend-beta is the label for that overlap.

It's common because it's easy to manufacture by accident. Backtest any long-biased system over a decade that happens to be mostly up, and the beta bleeds into the results whether or not you intended it. That's part of why walk-forward testing matters: a system that has only ever seen one direction of market has never had to prove it can do anything else.

In our audit trend-beta is the second most common rejection reason, trailing only no real edge itself. Trend-following systems were the single largest category we tested, 280 of them, and still failed 79% of the time, plenty of that driven by beta dressed up as a system. It's a different failure from tail-concentrated results, where the edge is real but fragile. Here there's no edge to be fragile. There's just a market that went up.

The uncomfortable part is that trend-beta isn't visible in the equity curve alone. A smooth, rising line looks identical whether it comes from genuine signal or from market drift, and a Sharpe ratio computed on that same rising instrument won't tell the two apart either. The only way to separate them is the benchmark comparison, done honestly, against the same instrument over the same stretch. Most of the time, that comparison is the whole answer.