Correlation trading uses the statistical relationship between cryptocurrency prices to construct positions that profit from relative strength shifts, mean-reversion between paired assets, or regime changes in market structure. In crypto markets, most altcoins move directionally with Bitcoin, but the degree of co-movement varies by market phase, liquidity conditions, and asset-specific catalysts. This guide covers how to measure crypto correlation, how BTC dominance functions as a regime signal, what altcoin beta means and how to calculate it, why correlations break down during stress events, how to apply correlation data to portfolio construction and pairs trading, and the common mistakes traders make when assuming stable relationships between assets.
What Correlation Means in Crypto Trading
Correlation measures how closely two assets move together over a defined period, expressed as a coefficient between -1 and +1. A coefficient of +1 means both assets move in the same direction by proportional amounts. A coefficient of -1 means they move in perfectly opposite directions. Zero means no linear relationship exists between their price movements.
In crypto, the Pearson correlation coefficient is the standard measure. It quantifies the strength of linear association between the log returns of two assets over a rolling window, typically 30, 60, or 90 days. The formula divides the covariance of two return series by the product of their individual standard deviations.
Practical ranges for crypto:
0.80 to 1.00: Near-lockstep movement. Holding both provides almost no diversification benefit.
0.50 to 0.80: Moderate positive correlation. Assets trend together but with meaningful variance in magnitude.
0.20 to 0.50: Weak positive correlation. Some directional overlap, but enough divergence to reduce portfolio risk.
-0.20 to 0.20: Negligible correlation. Movements are effectively independent.
Below -0.20: Negative correlation. Rare in crypto outside of stablecoin pairs or specific hedge structures.
Most large-cap altcoins maintain 30-day rolling correlations with BTC between 0.65 and 0.85 during normal market conditions. This number is not fixed. It shifts with market regime, narrative rotations, and liquidity events. Treating it as a constant is the first mistake correlation traders make.
The 30-day window captures recent behavior but can lag regime changes. Shorter windows (7-day) are noisier but more responsive. Longer windows (90-day) smooth out volatility but miss structural breaks. Professional correlation traders monitor multiple timeframes simultaneously and weight shorter windows more heavily when market conditions are shifting.
BTC Dominance as a Regime Signal
BTC dominance (BTC.D) measures Bitcoin's share of total cryptocurrency market capitalization. As of early 2026, it sits near 58%, indicating that more than half of all crypto market value resides in Bitcoin alone. This metric functions as a regime indicator rather than a trading signal.
Three regimes defined by BTC.D:
Rising dominance (BTC.D climbing above 55%): Capital is flowing into Bitcoin and out of altcoins on a relative basis. This does not require BTC price to rise. During the October-December 2025 correction, BTC fell from $126,000 to $67,000, but altcoins fell harder, with ETH dropping over 50% and SOL losing 60%. Dominance rose because Bitcoin declined less. In this regime, altcoin longs carry elevated risk, hedging with BTC shorts is less effective, and portfolio concentration toward BTC reduces drawdowns.
Falling dominance (BTC.D dropping below 54%): Capital rotates from Bitcoin into altcoins. Altcoins outperform BTC on a relative basis, often by large margins. Historically, sustained breaks below 54% dominance have preceded altcoin seasons where mid-cap tokens gain 200-500% while BTC moves 20-40%. Correlation between BTC and alts weakens during these rotations because altcoin-specific catalysts drive price rather than macro-crypto sentiment.
Stable dominance (BTC.D range-bound 54-58%): Neither regime dominates. Correlations tend to be moderate and stable. Pairs trades and relative-value strategies work best in this environment because the statistical relationships are most predictable.
How to read BTC.D shifts for trading:
BTC.D is not a timing tool. It identifies which playbook to use. When dominance is rising, reduce altcoin exposure and size positions smaller on alts. When dominance is falling, increase allocation to high-beta altcoins with strong narratives. When dominance is flat, relative-value and mean-reversion strategies between correlated pairs have the highest expected value.
Track BTC.D on weekly timeframes for regime identification. Daily fluctuations in dominance are noise. A sustained 3-4 week directional move of 2+ percentage points signals a genuine regime shift, not a one-day spike or dip.
Altcoin Beta: How Alts Amplify BTC Moves
Beta measures how much an asset's returns move relative to a benchmark. In crypto, the benchmark is typically BTC. An altcoin with a beta of 1.5 to BTC moves 1.5% for every 1% BTC moves, on average. Beta of 2.0 means the altcoin amplifies BTC moves by 2x in both directions.
Across our markets, we observe that altcoin correlation to BTC intensifies during selloffs and weakens during slow grinds higher, which means diversification benefits disappear precisely when traders need them most.
The beta formula:
Beta = Covariance(Alt Returns, BTC Returns) / Variance(BTC Returns)
This is a regression coefficient. You regress daily altcoin returns against daily BTC returns over a lookback period (typically 30-90 days). The slope of that regression line is beta.
Typical beta ranges (30-day rolling, 2025-2026 data):
Asset Category | Beta to BTC | Interpretation |
|---|---|---|
ETH | 1.0 - 1.3 | Slightly amplifies BTC moves |
Large-cap alts (SOL, ADA, AVAX) | 1.2 - 1.8 | Moderate amplification |
Mid-cap alts ($500M - $5B mcap) | 1.5 - 2.5 | Strong amplification |
Small-cap alts (under $500M) | 2.0 - 4.0+ | Extreme amplification, both directions |
Stablecoins (USDT, USDC) | ~0 | No meaningful beta |
What beta tells you for position sizing:
If you are long a 2.0-beta altcoin and BTC drops 5%, expect your altcoin to drop approximately 10%, plus or minus the residual (idiosyncratic) risk. This directly impacts your position sizing decisions. A 2% account risk on a 2.0-beta altcoin requires a position half the size of the same risk budget on BTC, because the expected move is twice as large.
Beta is asymmetric in practice. Many altcoins have higher downside beta than upside beta. They fall 2x when BTC drops but only rise 1.4x when BTC rallies. This asymmetry intensifies during high volatility environments and is a key reason why equal-weight altcoin portfolios underperform BTC over full market cycles despite occasional outperformance during rallies.
Calculating beta yourself:
1. Export daily closing prices for both assets over 60-90 days.
2. Calculate daily log returns: ln(Close_today / Close_yesterday).
3. Run a linear regression of alt returns (Y) against BTC returns (X).
4. The slope coefficient is beta. The R-squared tells you how much of the alt's movement is explained by BTC.
An R-squared of 0.70 means 70% of the altcoin's daily variance is attributable to BTC. The remaining 30% is idiosyncratic, driven by project-specific news, sector rotation, or liquidity changes. When R-squared drops below 0.40, the beta estimate becomes unreliable and the relationship may be breaking down.
Why Correlations Break During Stress Events
Crypto correlations are unstable. They increase during drawdowns and decrease during calm markets. This pattern, called correlation asymmetry, is the single most dangerous property for traders who assume diversification from holding multiple correlated assets.
The mechanism:
During sharp selloffs, forced liquidations hit leveraged positions across all assets simultaneously. Margin calls on BTC perpetuals trigger selling that cascades into altcoins as traders close entire portfolios to meet margin requirements. Cross-margined accounts liquidate every position at once, not just the losing one. This forced, indiscriminate selling pushes all crypto correlations toward 1.0 precisely when diversification is needed most.
Historical examples:
May 2021 crash: BTC fell 35% in a week. Correlations among top-50 assets spiked above 0.95. Every "diversified" crypto portfolio experienced near-identical drawdowns regardless of allocation.
November 2022 FTX collapse: BTC dropped 25%. SOL fell 60% (exchange-specific contagion). Correlation spiked initially, then SOL decorrelated to the downside as its specific risk dominated. This illustrates how correlation can spike and then break asymmetrically during idiosyncratic events.
October-December 2025 correction: BTC fell 47%. ETH fell 50%+, SOL 60%+, mid-caps 70-80%. The portfolio "diversification" from holding multiple alts provided zero protection because all correlations converged toward 1.0 under selling pressure.
What this means for risk management:
Never size positions based on normal-market correlations and assume that diversification will hold during drawdowns. Your risk of ruin calculation should assume correlations go to 1.0 in a crash. If your portfolio survives a scenario where every position drops simultaneously by its maximum historical stress-event loss, your sizing is appropriate. If it does not survive that scenario, you are over-allocated.
Correlation breakdown also creates opportunity. When a specific asset decorrelates from BTC due to an idiosyncratic event (protocol upgrade, regulatory ruling, hack), the divergence can be traded. But this requires identifying whether the decorrelation is temporary (tradeable mean-reversion) or permanent (structural shift requiring position exit).
Using Correlation for Portfolio Construction
Correlation data transforms portfolio construction from guesswork into a structured allocation process. The goal is not zero correlation between all holdings, which is nearly impossible in crypto, but rather minimizing the concentration of correlated risk.
Step 1: Map your correlation matrix.
Use a 60-day rolling correlation matrix of your held or watchlist assets against BTC and against each other. Tools like DefiLlama, Coin Metrics, or Sharpe Terminal generate these automatically. You want the full NxN matrix, not just each asset versus BTC.
Step 2: Identify correlation clusters.
Assets that correlate above 0.80 with each other form a cluster. Holding multiple assets within the same cluster does not diversify your risk. If SOL, AVAX, and NEAR all correlate at 0.85+ to each other, owning all three is economically similar to holding a larger position in one of them. Treat correlation clusters as a single risk unit when portfolio rebalancing.
Step 3: Allocate across clusters, not just across assets.
A portfolio with 5 assets in 2 clusters has 2 independent risk sources, not 5. Size each cluster's total allocation based on its aggregate beta and your risk budget. A high-beta cluster (large-cap alt layer-1s) deserves less total capital than a low-beta cluster (BTC + stablecoins) if your goal is controlled drawdown.
Step 4: Stress-test with correlation = 1.0.
Calculate your portfolio loss if all assets drop simultaneously at their individual 95th-percentile worst daily return. This is your crash scenario. If that number exceeds 15-20% of equity, reduce total exposure or add genuinely uncorrelated positions (stablecoins, cash, or short positions acting as hedges).
Step 5: Monitor for regime changes.
Rerun the correlation matrix weekly. When correlations shift by more than 0.15 over a 2-week period, the market regime is changing and your allocation assumptions may be stale. A sudden correlation drop between two previously locked-in assets can signal a trading opportunity or an emerging risk that your portfolio was not designed for.
For traders running delta-neutral strategies, correlation stability is a prerequisite. If the correlation between your long and short leg deteriorates, the hedge fails. Monitor the rolling 14-day correlation between legs and flatten the trade if it drops below your minimum threshold (typically 0.70 for crypto pairs).
Pairs Trading Basics in Crypto
Pairs trading profits from the relative movement between two correlated assets, going long the underperformer and short the outperformer when their price ratio deviates from its historical mean. The strategy is market-neutral in theory because you hold one long and one short position simultaneously.
The core requirements:
1. Two assets with historically high correlation (above 0.75 over 90+ days).
2. A stable mean ratio or spread between them.
3. A current deviation from that mean large enough to cover trading costs and generate profit.
4. A mechanism forcing reversion (arbitrage, fundamental linkage, or shared liquidity pool).
From an exchange operator's perspective, the pairs trades that generate the most margin-related support tickets are those where one leg gaps on an idiosyncratic event while the other barely moves, exposing the trader to asymmetric risk that no correlation model predicted.
Cointegration vs. correlation:
Correlation measures co-movement of returns. Cointegration measures whether the spread between two price series is stationary, meaning it fluctuates around a stable mean rather than drifting. Two assets can be highly correlated but not cointegrated if their ratio trends over time. For pairs trading, cointegration is the stronger requirement because it implies the spread will revert to its mean, not just that the assets move together.
Test for cointegration using the Augmented Dickey-Fuller (ADF) test on the spread series. A p-value below 0.05 suggests the spread is stationary and mean-reverting.
Common crypto pairs:
BTC/ETH: The most liquid pair. Their ratio oscillates within a range during stable regimes but can trend during altcoin seasons or Bitcoin-dominant phases.
SOL/AVAX: Both layer-1 smart contract platforms competing for similar capital flows. Ratio tends to mean-revert during stable market conditions.
ETH/BTC dominance-weighted: Long ETH when BTC.D peaks and starts declining, short ETH when BTC.D troughs and starts rising.
Execution mechanics:
Open both legs simultaneously to avoid directional exposure. Use perpetual contracts for the short leg since spot shorting is unavailable for most tokens. Size each leg so that a 1% move in the spread generates equal P&L on both sides. This means sizing inversely to price: if Asset A is $100 and Asset B is $50, you need twice as many units of B as A for dollar-neutral exposure.
Exit rules:
Exit when the spread reverts to its mean (profit target), when the spread widens further beyond a maximum threshold (stop-loss on the pair), or when the correlation between the two assets drops below 0.60 for 5+ consecutive days (relationship breakdown).
Pairs trading in crypto carries unique risks. Funding rates on the short perpetual leg can erode profits if held for days or weeks. Sudden delistings, hacks, or regulatory actions affecting one leg create asymmetric risk that no statistical model accounts for. Always use position limits that assume one leg can gap 20-30% overnight without the other moving proportionally.
Practical Tools for Tracking Correlation
Measuring and monitoring correlation requires consistent data and appropriate lookback windows. Manual calculation works for understanding the math, but active trading demands automated tools that update daily.
Free tools:
TradingView (BTC.D chart): Plots Bitcoin dominance in real-time. Use the correlation coefficient indicator overlay to compare any two assets on customizable lookback windows.
DefiLlama Correlation Matrix: Generates rolling correlation matrices for top crypto assets with configurable time windows. Good for identifying which assets cluster together.
CoinMetrics Charts: Research-grade correlation data with institutional methodology. Supports multi-asset comparison and downloadable datasets.
BitInfoCharts Correlations: Simple correlation table between major cryptocurrencies updated daily.
Paid/advanced tools:
Sharpe Terminal: Professional correlation matrices with 30d/90d/1Y/3Y windows. Includes cross-asset correlation with equities, bonds, and commodities for traders managing multi-asset portfolios.
Amberdata: API-level access to correlation and cointegration data. Supports backtesting pairs strategies against historical spread behavior.
Building your own tracker:
For traders who want custom lookback windows or asset combinations not available on free platforms: export daily close prices from any exchange API, calculate log returns, and compute the rolling Pearson coefficient in a spreadsheet or Python script. A 60-line Python script using pandas can generate a full NxN correlation matrix with rolling windows and flag threshold crossings. This is not complex engineering. It is basic data analysis that gives you an edge over traders relying on stale or default-window correlations.
What to monitor daily:
30-day rolling BTC/ETH correlation (baseline market structure signal).
BTC.D weekly trend (regime identification).
Any position-relevant pair correlations (for open pairs trades or hedges).
Sudden correlation drops (>0.15 in a week) as potential trade signals or risk flags.
Common Mistakes in Correlation Trading
Mistake 1: Assuming correlations are stable.
Crypto correlations shift faster than in traditional markets because the asset class is young, liquidity profiles change rapidly, and narrative-driven capital flows can override fundamental relationships. A correlation that held at 0.85 for six months can drop to 0.40 in two weeks if one asset gets a major protocol upgrade or regulatory ruling. Never build a strategy on the assumption that today's correlation will hold next month.
Mistake 2: Using a single lookback window.
A 90-day correlation might show 0.80, but the last 14 days might be 0.50 because the relationship is breaking down. Using only one window blinds you to regime changes. Always compare short (7-14 day), medium (30-60 day), and long (90-180 day) windows. When they diverge significantly, the relationship is in transition.
Mistake 3: Confusing correlation with causation.
Two altcoins being correlated does not mean one causes the other to move. Both might be responding to Bitcoin, which drives 65-70% of altcoin variance. If BTC is the common driver, the pair will only mean-revert when BTC influence is stable. If BTC regime changes, the pair relationship changes too.
Mistake 4: Ignoring correlation asymmetry.
Correlations during rallies differ from correlations during crashes. Most traders measure correlation across all conditions and get an average that is misleading. Calculate upside correlation (returns on days BTC is positive) and downside correlation (returns on days BTC is negative) separately. The downside number is what matters for risk management. It is almost always higher.
Mistake 5: Over-diversifying within a single cluster.
Holding 8 layer-1 altcoins that all correlate at 0.85+ to each other provides the illusion of diversification. In reality, you hold one concentrated bet with 8 different names. This mistake compounds during drawdowns when all 8 positions decline together and aggregate losses exceed what the trader expected from a "diversified" portfolio. Treat market cycles as the dominant driver and cluster accordingly.
Mistake 6: Neglecting transaction costs in pairs trades.
A pairs trade requires opening two positions and closing two positions. With taker fees, slippage, and potentially funding rates on the short leg, your round-trip cost can easily reach 0.3-0.5% of notional. If the expected spread reversion is only 2%, your costs consume 15-25% of the profit. Only trade spreads where the expected reversion is at least 3-4x your all-in transaction cost.
Frequently Asked Questions
What is a good correlation threshold for pairs trading in crypto?
A minimum 90-day Pearson correlation of 0.75 between the two assets is the starting threshold. Below that, the relationship is too weak for reliable mean-reversion. Above 0.85, the spread may be too tight to trade profitably after costs. The ideal range is 0.75-0.90 with confirmed cointegration via an ADF test (p-value below 0.05). Additionally, check that the 30-day correlation has not diverged more than 0.15 from the 90-day number, which would indicate an unstable relationship.
How often should I recalculate crypto correlations?
For active traders, recalculate weekly using a 30-day rolling window and compare against a 90-day baseline. If any pair in your portfolio shows a divergence greater than 0.15 between the two windows, investigate whether a regime change is underway. For portfolio-level allocation decisions, monthly recalculation with a 60-day window is sufficient. Never set a correlation assumption once and forget it for months.
Does BTC dominance predict altcoin prices?
BTC dominance does not predict absolute price direction for altcoins. It signals relative performance. Falling dominance means altcoins are outperforming BTC, but both can be falling in dollar terms (altcoins just falling less). Rising dominance means Bitcoin is outperforming, which can happen while both BTC and alts rise (BTC just rising more). Use dominance as a regime filter to select which strategy playbook to run, not as a directional price predictor.
Can correlation trading work with small accounts?
Yes, but the minimum practical account size is higher than for directional trading because you need to hold two positions simultaneously and pay fees on both legs. On BloFin perpetuals, a small pairs trade requires enough margin for both a long and short position plus a buffer for adverse movement. Practically, $2,000-$5,000 is the minimum to run a single pairs trade with appropriate spot hedging and survive a temporary spread widening without hitting margin limits.
What causes two previously correlated crypto assets to decorrelate?
The most common causes are: (1) one asset receives a major protocol upgrade or exploit that changes its fundamental value proposition; (2) regulatory action targets one asset specifically; (3) one asset gets adopted by a new user base (institutional inflows into ETH ETFs, for example) that trades on different signals than the crypto-native market; (4) liquidity dries up in one asset while remaining stable in the other, causing the thinner asset to gap on small orders; (5) sector rotation where capital exits one narrative (layer-1s) and enters another (AI tokens) without affecting BTC.
Researched and written by the Blofin Academy editorial team with AI-assisted drafting. Primary sources include BloFin exchange documentation (perpetual contract specifications, margin modes, fee schedules); CoinMetrics correlation data and methodology (Coinmetrics, https://coinmetrics.io/); TradingView BTC.D dominance chart (TradingView, https://www.tradingview.com/symbols/BTC.D/); Amberdata pairs trading research on cointegration vs. correlation (Amberdata, https://blog.amberdata.io/); DefiLlama correlation matrix (DefiLlama, https://defillama.com/correlation). All facts independently verified against cited documentation current as of April 2026.
This article is for informational purposes only and does not constitute financial advice. Cryptocurrency trading involves substantial risk of loss. Past performance does not guarantee future results. Always conduct your own research and consider your financial situation before trading. BloFin does not guarantee the accuracy of third-party data referenced herein.