Σ
Dimensionless.quant
Back to Strategies
REJECTEDCrypto Arbitrage Research Lead · Apr 02, 2026

Crypto Funding Reversal

Mean-reversion strategy trading perpetual futures funding rate extremes across major crypto exchanges (Binance, Bybit, Deribit).

Strategy Score
34

Edge concentrated in 2021-22 funding spikes. No persistence post-2023. Out-of-sample Sharpe 0.21.

Performance Metrics
WF Sharpe
1.94
OOS Sharpe
0.21
DSR
0.08
Regimes
1/5

How It Works

Fade extreme funding rates by taking opposite positions when funding reaches historical percentiles, expecting reversion to neutral.

Mechanics

  1. 1.Monitor perpetual funding rates for BTC, ETH across 3 exchanges
  2. 2.Enter short perp when funding rate > 95th percentile (e.g., +0.1% 8hr)
  3. 3.Enter long perp when funding rate < 5th percentile
  4. 4.Hold position until funding reverts to 50th percentile or 7-day timeout
  5. 5.Hedge spot exposure with inverse perp on backup exchange

Signals

Funding rate percentile rank (rolling 90-day window)Open interest changes (confirms positioning extremes)Spot-perp basis spread

Performance Results

Sharpe Ratio
Walk-Forward1.94
Out-of-Sample0.21
In-Sample3.12
Returns
Annualized2.8%
Max Drawdown-31.4%
Calmar Ratio0.09
Consistency
Win Rate48%
Profit Factor1.1
Avg Win/Loss1.8% / -2.1%
Capacity Analysis
Max Capacity
N/A
Current Slippage
N/A
At Capacity
N/A

Implementation

Instruments

BTC-PERP, ETH-PERP on Binance, Bybit, Deribit

Execution

Aggressive limit orders, 5-minute TWAP for larger sizes

Rebalancing

Event-driven (triggered by funding extremes)

Risk Limits

Max 2% portfolio allocation, -5% stop loss per trade

Technology

WebSocket feeds, Python execution, CCXT library for exchange APIs

Risk Analysis

Funding Persistence

High Impact

7-day timeout prevents indefinite holding

Exchange Risk

High Impact

Diversify across venues, but insufficient in 2023 blowups

Regime Shift

Critical Impact

None effective - post-2022 funding volatility collapsed

Execution

Medium Impact

Limit orders, avoid market orders in thin books

Backtest Results

Period
Jan 2020 - Mar 2026
Total Trades
1,243
Avg Holding
3.2 days
Best / Worst Year
+48.2% (2021) / -18.9% (2024)

Stress Period Analysis

2021 Bull Run+48.2%

Funding spikes frequent, strong edge

2022 Bear Market+22.1%

Negative funding extremes profitable

2023-2026 Normalization-12.4%

Funding volatility collapsed, edge disappeared

Deployment Status

Status

Rejected - not deployed

Rejection Analysis

Post-2022, crypto perpetual funding markets normalized. Funding rate extremes became rare and short-lived. Out-of-sample testing (2023-2026) showed Sharpe of 0.21, well below deployment threshold. Edge appears structural to 2021-22 mania period and not persistent. Risk Committee rejected deployment due to regime dependency and lack of recent performance.