Crypto Correlation Trading: Entries, Hedges, Exits

Crypto correlation trading works best when you stop asking which coin is bullish and start asking whether one coin is mispriced against another. I use it for BTC/ETH, ETH/SOL and large-cap alt spreads when direction is noisy but relative strength is clean.

This is not a magic hedge. It is a structured way to trade mean reversion, rotation and temporary dislocations while keeping less naked market beta than a simple long or short.

When does crypto correlation trading actually give an edge?

The edge appears when two assets normally move together, then temporarily separate because of positioning, funding, news flow or liquidation pressure. Among crypto trading types, this sits between discretionary relative value and simple statistical arbitrage.

For liquid pairs, I want 30-day correlation above 0.75 before I care. Then I wait for the spread to move at least 2 standard deviations from its 30-day average before entering.

VoiceOfChain tracks live correlation, spread z-scores and funding divergence across Binance, Bybit and OKX — you can see BTC/ETH and ETH/SOL dislocations without building the sheet yourself. https://voiceofchain.com

How do I measure correlation fast in TradingView?

The quickest tradingview crypto correlation setup is the Correlation Coefficient indicator. Put BTCUSDT from Binance or Coinbase BTCUSD on the chart, add the indicator, then compare it against ETHUSDT, SOLUSDT or another liquid pair.

I use 30-day and 90-day windows. The 30-day window tells me whether the trade is active now; the 90-day window tells me whether the relationship is structurally reliable.

• Use the same quote currency where possible: BTCUSDT vs ETHUSDT is cleaner than BTCUSD vs ETHBTC.
• Use exchange pairs with real liquidity: Binance, Coinbase, Bybit and OKX are usually cleaner than thin alt venues.
• Ignore one-candle correlation spikes after CPI, FOMC or liquidation cascades; they often fade within hours.
• Do not rely on generic crypto trading reviews of tools; check whether the tool shows window length, symbol source and spread z-score.

import pandas as pd

# df has columns: btc_close, eth_close
returns = df[['btc_close', 'eth_close']].pct_change()
df['corr_30d'] = returns['btc_close'].rolling(30).corr(returns['eth_close'])
spread = (df['eth_close'].pct_change(30) - df['btc_close'].pct_change(30))
df['zscore'] = (spread - spread.rolling(30).mean()) / spread.rolling(30).std()
print(df[['corr_30d', 'zscore']].tail())

What entry rules do I use for BTC/ETH or ETH/SOL pairs?

My basic crypto correlation trading strategy is simple: only trade liquid pairs, only enter after a measurable stretch, and exit when the spread normalizes. If BTC is at $63,200 and ETH is at $1,625, I am not predicting both charts; I am trading ETH's relative move against BTC.

• Rule 1: 30-day correlation must be at least 0.75 and 7-day correlation must stay above 0.60.
• Rule 2: Enter only when the spread reaches +2.0 or -2.0 z-score.
• Rule 3: If ETH underperforms BTC by -2.0 z-score, long ETH and short BTC.
• Rule 4: If ETH outperforms BTC by +2.0 z-score, short ETH and long BTC.
• Rule 5: Exit half at 0.8 z-score and the rest between 0.2 and 0.0.
• Rule 6: Kill the trade if 7-day correlation drops below 0.45.

I do not enter right after a coin-specific event like an ETH ETF decision, a Solana outage, a major token unlock or a Binance delisting notice. Those are correlation breakers, not mean-reversion setups.

How should I size the hedge and place stops?

Start with risk, not leverage. On a $20,000 account, I usually risk 1% on a correlation trade, so the maximum planned loss is $200.

Assume ETH beta to BTC is 1.25. To reduce BTC market exposure, $8,000 long ETH hedges against roughly $10,000 short BTC because 8,000 multiplied by 1.25 equals 10,000.

If the spread stop is estimated at a 2.5% loss on gross exposure, $8,000 gross risks about $200. If the target mean reversion is 4.5%, the expected profit is about $360, giving a 1.8R setup.

• Use a spread stop first: exit if z-score moves from -2.1 to -3.0 instead of mean reverting.
• Use a correlation stop second: exit if 7-day correlation closes below 0.45.
• Use a funding stop third: exit or reduce if carry cost exceeds 0.10% per 8-hour window against your long leg.
• Keep liquidation far away: I rarely use more than 2x to 3x effective leverage on pair trades.

What can go wrong when the correlation breaks?

The most common mistake is treating correlation like a promise. It is only a recent relationship, and crypto relationships can break violently when one asset gets a fresh catalyst.

Funding can also turn a correct trade into a bad trade. On Bybit perpetuals, if you are long a leg paying +0.10% every 8 hours on $10,000 notional, that is $10 per interval, $30 per day and $210 over a week.

• Coin-specific news: ETH upgrade, SOL outage, token unlock or SEC headline can invalidate the spread.
• Funding drag: Binance, Bybit and OKX contracts can have different funding intervals or caps, so check each pair before entry.
• Liquidity mismatch: Gate.io or KuCoin alt perps can wick 5% to 8% while BTC barely moves.
• Liquidation cascades: a crowded long-short spread can unwind faster than the correlation model updates.
• Wrong hedge ratio: dollar-neutral is not always beta-neutral, especially with high-beta alts.

The honest risk caveat: this approach fails during regime shifts. When the market starts repricing one asset for a real reason, averaging down on correlation is just a slower way to get run over.

Frequently Asked Questions

The key takeaway: crypto correlation trading is a spread trade, not a prediction engine. The money is made by waiting for a strong relationship to stretch, entering with a defined hedge ratio, and exiting when the spread normalizes.

Use hard rules for correlation, z-score, funding and position size before you place the order. If the relationship breaks, take the stop and move on; the next clean dislocation is always better than defending a broken pair.