Open
Portfolio Regime Research with RLX
How to find profitable market entry regimes across multiple strategies and turn them into reusable allowlists.
Research library · human and agent studies
The write-upsbehind every verdict.
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.
SurvivedBTCUSDT · 1h
Autopilot Strategy Creation & Evolution with AI Agents and Deep RL
A guide on using the RLXBT Deep Reinforcement Learning (DQN) engine and MCP integrations to train, evaluate, and predict market signals dynamically using AI agents.
- Total return
- +0.16%
- Sharpe
- 2.34
- Max drawdown
- 0.06%
Open
Optimizing RL Agents with Exit Rules in RLXBT
In this article, we explore how using strict Exit Rules affects the training and performance of Reinforcement Learning (RL) agents in the cryptocurrency market.
SurvivedBTCUSDT · 1h
Robust RSI + Bollinger Mean Reversion Strategy on BTCUSDT 1h
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.
- Total return
- +49.16%
- Sharpe
- 0.75
- Max drawdown
- 15.35%
OpenBTCUSDT · 1h
Squeeze-Release Short: Volatility Squeezes, Information Entropy, and Parameter Decay on BTCUSDT 1h
An in-depth quantitative analysis of the Custom Squeeze-Release Short strategy on BTCUSDT 1h, exploring volatility compression, Shannon entropy, return z-scores, and walk-forward parameter decay.
- Total return
- +10.90%
- Sharpe
- 1.64
- Max drawdown
- 2.62%
Failed18 cryptocurrency futures WML · 1w
The Momentum Mirage: +106% One Year, -43% the Next
Weekly crypto momentum survived fees and funding, gained 106% in 2025, then reversed to -43% in the untouched 2026 test. Volatility scaling did not save it.
- Total return
- -43.24%
- Sharpe
- -2.33
- Max drawdown
- 42.12%
FailedBTC-led 16-altcoin futures quintiles · 5m
Sharpe 10.73, Still Untradeable: The Cross-Crypto Experiment Gets 16 More Chances
A 16-altcoin BTC-led portfolio achieved gross Sharpe 10.73—but turnover costs were still 26.6 times larger than its alpha.
- Total return
- -99.97%
- Sharpe
- -208.47
- Max drawdown
- 99.96%
SurvivedBTCUSDT · 1h
Curing the Falling Knife: High-Efficiency Trend Pullbacks with Limit Order Absorption
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.
- Total return
- +0.16%
- Sharpe
- 0.96
- Max drawdown
- 0.10%
SurvivedBTCUSDT · 1h
BTC 4-Leg Combo Strategy: Sharpe 1.02, +294% Return Under 0.05% Commissions
The definitive multivariable combo strategy on BTCUSDT 1h, out-of-sample validated under realistic commissions.
- Total return
- +294.30%
- Sharpe
- 1.02
- Max drawdown
- 29.90%
SurvivedBTCUSDT · 1h
Curing Breakout Parameter Decay: Symmetrical Volatility Squeezes with Shannon Entropy Filters
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.
- Total return
- +0.07%
- Sharpe
- 0.47
- Max drawdown
- 0.12%
Failed
The In-Sample Trap: A Bollinger-Band Mean-Reversion That Dies Out-of-Sample (BTCUSDT 1h)
A textbook Bollinger-band fade looks tradeable in-sample - positive Sharpe, ~50% win rate over 1,291 trades. Walk-forward analysis kills it. A short case study in why out-of-sample validation is non-negotiable.
- Sharpe
- 0.37
- Trades
- 1,291
- Win rate
- 50.40%
SurvivedBTCUSDT · 1h
BTC 1H Hurst Breakout: Fractal-Filtered Trend Trading Under Commissions
A robust volatility breakout strategy on BTCUSDT 1h, gated by Hurst Exponent and Fractal Dimension Index, out-of-sample validated under real fees.
- Total return
- +19.13%
- Sharpe
- 0.72
- Max drawdown
- 6.50%
OpenBTCUSDT · 1h
Custom Squeeze-Release Short Strategy with Custom Features on BTCUSDT 1h
Short-only strategy based on custom squeeze-release indicators on BTCUSDT. Backtest shows Sharpe 1.64 and low drawdown, but WFE of 0.35 signals high potential for overfitting.
- Total return
- +10.90%
- Sharpe
- 1.64
- Max drawdown
- 2.62%
Survived
Optimized BTC 5m Oscillation Exhaustion: High-Frequency Bottom Reversion Slashed to Clear Maker-Only Trading Fees
Why a 'modest' 1.5% unleveraged return over 9 months is actually a major quant victory. We break down capital efficiency (1.67% time-in-market), leverage scaling, and overcoming transaction friction.
- Total return
- +1.55%
- Sharpe
- 0.43
- Trades
- 110
OpenBTC · 1h
I Asked an AI Agent to Find a Tradable Candlestick Pattern. It Refused to Pretend Three Trades Were Proof.
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.
- Total return
- +4.16%
- Sharpe
- 1.94
- Max drawdown
- 1.99%
Open
A Cross-Asset Signal That Survived Costs — and Still Isn't Tradable Yet
BTC-crash → ETH/SOL cross-asset signal: Sharpe 0.96 on 96 events, WFE 0.47, costs eat 62% of edge. WATCH verdict. 69-day corpus, no sealed OOS.
OpenETHUSDT · 1h
Comprehensive Price Action Study: Systematic Short Bias and Failure of Bullish Hammers
A systematic study of 7 classic Price Action patterns across BTC, ETH, and SOL. Reveals a strong SHORT bias, the failure of Bullish Pin Bars, and a tradable contrarian fading edge (+5.59% ETH 1h).
- Total return
- +5.59%
- Sharpe
- 0.33
- Max drawdown
- 1.72%
SurvivedAPI · manifest
RLXBT Agent Ingestion Guide: Enabling Autonomous AI Agents to Replicate Strategies
A programmatic guide for autonomous quant agents on how to parse, extract, and replicate our strategies using the RLXBT API and local daemon toolchains.
- Total return
- +0.00%
- Sharpe
- 0.00
- Max drawdown
- 0.00%