Research
What actually survives real costs
Cost-aware findings from our audit of 1,000+ TradingView strategies and 1,700+ indicators — over 2,700 scripts in all. No hype, no signals. What held up under real spreads, commissions and slippage, and what quietly fell apart once the cost model stopped being a guess.
Every strategy is ported to our own Python engine and tested on 13 years of CME futures, tick-level FX with real bid/ask, liquidity-aware stocks and crypto. The survivors then face a second, adversarial pass that hunts for look-ahead bias. We reject about 78% of the strategies — and the biggest reason isn't cost. Most never had an edge to lose. These are the findings, broken down by strategy type and by how they fail.
The Academic Audit
We re-ran the famous published quant factors — Betting-Against-Beta, low-volatility, the 52-week-high effect, book-to-market value — on a survivorship-free, cost-aware panel, each linked to its paper.
48 of 54 failed after real costs · only 4 survived, as market-neutral factor-legs
The big pictureWhy most trading strategies fail
The single most common reason a strategy gets rejected — and it isn't trading costs. Most never had an edge to begin with.
~78% of strategies rejected · no edge is the #1 cause
Strategy typeDo grid & DCA bots actually work?
We ran 70+ grid, DCA and martingale bots through the audit. The result was unusually clean.
100% failed — the only category with zero survivors
Strategy typeDo trend-following strategies work?
The biggest category we tested. Most of the winners turned out to be market drift wearing a system's clothes.
~79% of 280+ rejected — roughly the house average
Strategy typeDo mean-reversion strategies work?
They print steady wins in a range, then give it all back in the first real trend. We measured how often.
~80% of 90+ failed — win in ranges, die in trends
Strategy typeIs momentum trading profitable?
One of the least-bad categories in the whole audit — which still means most of it did not survive.
~73% of 140+ rejected
Strategy typeDoes ICT / SMC actually work?
Smart Money Concepts tested the same cost-aware way as everything else. Better than its reputation suggests — and still mostly failed.
~73% of 20+ rejected
New findings are published regularly — by strategy type, by failure mode, and what the numbers actually say.