We test crypto strategies with academic rigor. When they don't work, we publish that too.
Most crypto strategies look profitable — until you test them on data they've never seen. We run that test on every idea, in public, and document the result whether it survives or dies.
143,535 funding settlements · 2022–2026 · as of May 2026
Current regime, recalculated from market data
Our research infrastructure classifies the structural state of the crypto market — trend and volatility — from real price history. ApexQuant classifies market structure. It does not predict returns.
The strategies we tested and rejected
Every failed hypothesis, documented publicly with full IS/OOS results. Nobody else in crypto does this.
Now testing
Hypothesis #5 — spec frozen before any code.
Verdict published when out-of-sample completes: survives or joins the graveyard.
Volatility Breakout
“Trend capture on volatility expansion”
Postmortem in preparation
Single-asset
In preparationCross-Sectional Reversion
“Mean reversion across the cross-section”
Postmortem in preparation
Cross-sectional
In preparationIndependent quantitative research
Original analysis, academic paper replications, and methodology. Every claim includes the data and code that produced it.
How we test honestly
The crypto research industry has a survivorship bias problem. Our method is built to remove it.
Strict IS/OOS splits
We never optimize on the same data we report. Out-of-sample is frozen and touched exactly once.
A priori verdicts
Pass/fail thresholds are defined before we see any results. No moving goalposts after the fact.
No parameter mining
One parameter choice, justified in advance. We don't sweep fifty thresholds until one looks good.
Public Strategy Graveyard
Every failed hypothesis is documented with full results. Negative findings are first-class output.
Data, validated before it ships
A small set of free, daily-updated views. The rest is reserved for subscribers — and only published once it passes our own validation.
Cross-asset funding & carry, ranked by structural quality.
Persistence and autocorrelation scoring per symbol.
BTC return-regime classification, updated daily.
Cross-sectional funding rate spread across 24 pairs.
Basis and funding divergence across venues.
Full dashboards open as the research lab validates each layer.
See what's included →Be there when the first strategy survives.
So far: 0 of 4 made it past out-of-sample. Every verdict lands in your inbox — survivors and casualties alike. Free, no spam, unsubscribe anytime.
Join researchers and allocators who prefer method over hype.