The promise: we show the whole scoreboard
Most platforms show you the strategies that “work” and quietly bury the rest. We do the opposite. Here is the complete, unedited verdict from months of exhaustive testing — the failures in full, and the one thing that survived. If you only read one thing about crypto strategy research, read this.
The single test that decides everything is the beta-strip: subtract the market’s own return. If a strategy’s edge disappears once you remove the market move, it was never an edge — it was just market exposure (beta) wearing a costume.
What failed (and it’s almost everything)
| What we tested | Result |
|---|---|
| 14 indicator strategies (momentum, MA, RSI, MACD, Bollinger, Ichimoku, …) | 0 of 14 — all beta |
| Trendline trading (116,400 trades) | 39% win rate — beta/noise |
| Triangle / consensus / multi-signal voting | beta |
| Bollinger + Ichimoku combinations (6 variants) | 0 of 6 |
| Indicator + filter × timeframe (36 cells) | 0 of 36 |
| Cross-sectional momentum / reversal / low-volume | artifacts (die in liquid universe) |
| Low-volatility factor | disguised bear-beta (leg-decomposed) |
| Altcoin tiers (capacity-aware) | worse — cost eats it |
| Every exit method (fixed, trailing, signal-reverse) | conserves expectancy, doesn’t create it |
| Non-linear / machine learning (GBM) | negative even in-sample (no free lunch) |
The pattern is not bad luck. It is structural: for retail, on public data, directional prediction has an edge smaller than trading costs. You can dial a strategy’s win rate anywhere from 21% to 87% by changing the exit — and the expectancy stays negative the whole time. The edge has to come from the entry, and directional entries don’t have one.
What survived
One thing passed the beta-strip: the volatility risk premium — selling volatility, where option buyers systematically overpay for protection and the seller collects the difference. It does not predict direction; it harvests an insurance premium. Correlation to the market: near zero. That is a real risk premium, not disguised beta.
We are honest about its status: it is a strong lead, not yet live. It trades on OKX ETH options (small enough for a real account, no other venue needed), and it goes live only after a real-option-chain backtest passes around 2026-08. Here’s exactly what we’re validating, in the open.
Why we publish the failures
Because the failures are the product. A platform that only shows winners is indistinguishable from a scam — and most are scams. The honest scoreboard, including every loss and why it lost, is the one thing you cannot get from a guru selling signals.
Test any of these yourself on the simulator — with real fees, across real coins. You will reach the same verdict we did. That is the point: don’t believe us. Verify.