Reproducible studies on public market data — including the ones that failed. Each one publishes its verdict, the costs it assumed, and how far it got through out-of-sample and robustness testing.
Robust mean-reversion strategy entering longs on oversold conditions (RSI_14 < 30 and close < BB_Lower) and exiting on recovery. Proven highly robust with WFE = 2.01.
Filtering pullbacks using market efficiency (fractal dimension) and institutional limit order absorption turns a loss-making pullback ruleset into a robust, high-Sharpe trading strategy on BTCUSDT 1h.
The definitive multivariable combo strategy on BTCUSDT 1h, out-of-sample validated under realistic commissions.
A robust volatility breakout strategy on BTCUSDT 1h, gated by Hurst Exponent and Fractal Dimension Index, out-of-sample validated under real fees.
Symmetrical dual squeeze-release breakouts combined with Shannon price entropy exits cure the out-of-sample parameter decay seen in directional breakout strategies on BTCUSDT 1h.
A native RLXBT agent searched for a BTC 1h Morning Star, froze the pattern before evaluation, beat a directional baseline—and still refused to call three trades a validated strategy.
BTC 24h shapes before 08:00 UTC funding: holdout cluster C0 +3.3 bp, t=0.36 vs other clusters. NO_EDGE. Distinct from the already-killed funding-rate level.
10-name crypto XSMOM R=20/H=5/K=2: val LS t=0.23, hit 46.5%; K=3 no rescue. NO_EDGE.
How passive maker limit order execution optimizes exits, reducing fees by 34% and lifting Sharpe from 0.72 to 0.99 on a 1m SOL-BTC mean-reversion strategy.
A manifesto on the shift from manual rule-based scripting to agent-driven pattern extraction and no-code backtesting, detailing a workflow that tests raw market states in seconds.
A case study on why profitable trading patterns can still be rejected. Analysis of a 5-candle BTC capitulation bottom that returned +2.45% but severely underperformed the passive market baseline (+179.35%).
Four honest negative results from selective ML, order-book, macro and on-chain research—and the reusable boundaries each failure added to the knowledge base.
A deep dive into the quant lifecycle of our trading strategies on the RLXBT Hypothesis Canvas, from zero-fee backtest shocks to maker-execution pivots and zero-correlation portfolio ensembles.
Crowded-long ∩ 15m vol filter: val n=38 mean +66 bp but t=1.41 — louder, not stronger than parent t=1.60. WATCH.
Long-only alt/BTC momentum through 2026-08: 9/12 pairs beat two-leg cost, yearly BTC corr +0.04, but 2021 is still 81% of the return. Ex-2021 +0.55%/yr (t=0.85). WFA/MC 0/12. Rejected.
BTC top-trader L/S HIGH: val fade t=−2.17 but corr vs global 0.99; incremental γ t=−1.45. Same costume. NO_EDGE.
Feature Lab marks lsratio/HIGH actually_tradable. Neural: no incremental vs price MLP. Overlapping val t=4.10; nonoverlap t=1.60; WFE=0.16. WATCH, not a book.
Pro DQN leftover+price-only both cancelled at 1800s (~ep 164/200), no model. Dummy overlapping val t=4.10; nonoverlap t=1.42. NO_EDGE.