[ LOADING… ]
We have tested eight trading strategies the same honest way. Five we graded as no edge. Three we could not conclude either way. Zero have a confirmed edge on unseen data. This page shows the work, including the dead ends. Judge for yourself.
Crypto signal-sellers post curated wins, crop the losers out of the screenshot, and never tell you how the strategy was tested — or whether it was tested at all. A backtest you cannot see and cannot reproduce is not evidence. It is a sales asset.
The thing that is actually broken in this market is not signals. It is trust. So we built the opposite: every strategy runs through one validation standard, the results are published whether they are good or bad, and what we tell you matches what is true by construction. Most of what follows is us failing to find an edge. That is the point — you are seeing the real filter, not the highlight reel.
Two rules decide whether a strategy is allowed anywhere near real money.
We tune a strategy's settings on one slice of history, then score it on a later slice it has never touched — walking that window forward through years of data. A strategy that only looks good on the data it was fitted to gets caught here. Live results have to survive data the strategy never saw.
Test enough variations and one will look brilliant by pure luck. So we raise the pass mark in proportion to how many things we tried — a deflated Sharpe ratio. On our program the honest bar works out to roughly 2.94. Nothing we have tested has cleared it.
We also charge every trade a realistic 0.30% round-trip cost, because a strategy that only wins before fees does not win. That single number has killed more of our ideas than any other.
Same open-source research harness (quantcore, walk-forward v0.3.0) and the same code path for all eight strategies — no special cases.
Every strategy we have put through the harness, with its honest verdict. “No edge” means it was flat-to-losing on data it had not seen. “Inconclusive” means too few trades to call it either way — not a quiet win we are hiding.
Research results as of 2026-06-09 · quantcore walk-forward v0.3.0. These are static research verdicts, not live numbers.
Losing and statistically significant on all 8 coins. This was our live flagship.
Actively destructive — lost on every coin, in both directions.
A significant loser; the move it bet against simply kept going.
Flat-to-negative — a second momentum idea, same dead end.
Too few trades to test fairly — not a quiet win we are hiding.
Starved by limited funding-rate history; untestable, not killed.
Turtle (daily, long-only) is the only strategy that leaned positive at all — a real-looking tilt in long-side trend-following. We are saying plainly: it did not pass. The sample is too small, and once we corrected for trying multiple settings, most of the apparent strength was luck-of-the-fit. It is the one idea worth feeding more data — not a green light, and not something we would dress up as proven.
The honest takeaway: across these eight strategies, on these coins, at these timeframes, after real costs — we found no edge in either trending or fading the market. That is a negative result. We think it is the most valuable thing we own, because it is the proof that our standard is real and that we will not sell you something that did not pass it.
Turtle is the one result with a pulse, so we are running it live on a small scale to gather the out-of-sample evidence the backtest could not. This is a forward test, not a proven strategy. You can watch it arm and wait for a breakout in real time. If it fails live, you will see that too.
Our old flagship graded out as no edge. We keep a small version running and fully visible specifically because we graded it negative — so the track record includes a strategy we already told you does not work, rather than quietly retiring the inconvenient one.
Neither of these is a “winner.” One is an experiment; the other is a documented dead end we are being honest about.
We do not have a proven, profitable strategy today, and we will not pretend otherwise. What we have is a standard strict enough to reject our own ideas, the discipline to publish the rejections, and real capital on the line so the results are honest.
We believe any edge we eventually find will be small, conditional, and capacity-limited — the kind big money will not bother with — and it will have to clear the same bar everything above failed. You will see every step of that hunt, dead ends included.
Do not take our word for any of it. Every trade we place — wins and losses, from real money — is on the live track record.