- BTC ·
ETH · SOL · XRPAssets - Shared clockAligned
- Rolling correlationMeasure
- 5m ·
15m · 1h · 1dTimeframes
When BTC moves, the rest of crypto usually follows, but not always, and not equally. Because the four coins’ Up/Down markets are captured on the same clock, you can measure how tightly their implied odds move together, and spot the moments when that link breaks down.
Correlation is only meaningful when the series are aligned in time. Since BTC, ETH, SOL, and XRP snapshots share the same capture pipeline and millisecond stamping, lining up their mid-price series is clean, no resampling guesswork, which makes the correlation you compute trustworthy.
What correlation tells you
Co-movement regime
High, stable correlation means the coins are trading as one risk asset, a market-wide move, not a coin story.
- Tight co-movement
- Macro-driven
- Diversification is low
Breakdowns
When one coin decouples, correlation drops, often a coin-specific catalyst the others have not priced.
- Decoupling events
- Relative-value setups
- A coin going its own way
Lead and lag
BTC often leads. Measuring which coin moves first turns one market into an early read on the others.
- BTC as leader
- Lagging followers
- Early directional read
Strategies it opens up
- Rotation, lean toward whichever market is leading a co-ordinated move, rather than spreading thin across all four.
- Relative value, when a usually-correlated pair diverges, fade the gap if you expect the link to reassert.
- Risk awareness, in a high-correlation regime, four crypto positions are closer to one bet than four; size accordingly.
- Timeframe nuance, correlations look different at 5-minute and daily horizons; pick the one your strategy trades.
Correlation is not a fixed property of these coins, it is a regime that tightens and loosens. The value is in measuring the change, not memorising a single number.
How to build the correlation series
The raw material is the mid-price of each coin’s Up/Down market, the implied probability between 0 and 1, sampled at whatever timeframe matches your horizon. Because every crypto snapshot carries the same set of fields, the workflow is identical across all four assets: pull the mid-price series, align on the shared timestamps, then roll a window across them.
- 1Pick a timeframe (5m for intraday work, 1h or 1d for regime context) and pull the matching Up/Down snapshots for each coin.
- 2Use mid_price, the implied probability, as the comparable series; it already nets best bid and best ask.
- 3Align the four series on their millisecond timestamps; no interpolation is needed because they share one capture clock.
- 4Compute a rolling correlation over a window sized to your horizon, and watch where it spikes or collapses.
- 5Re-run across resolved markets too, so the picture reflects a population of days rather than a single live snapshot.
Same clock, honest correlation
A correlation computed on mismatched timestamps is noise dressed as signal. Because all four coins are captured and stamped by the same pipeline, their series line up exactly, so the rolling correlation you measure reflects the markets, not a resampling artefact.
What a correlation regime looks like
- High & stableCoins trade as one risk asset, macro regime
- FallingA coin is decoupling, relative-value window opening
- NegativeRare, usually a sharp coin-specific catalyst
- ChoppyNo clean regime, lower conviction, smaller size
Treat these as research lenses, not signals with guarantees. A falling correlation tells you a relationship is changing; it does not tell you the gap will close on your timeframe. The discipline is to measure the regime, size to your confidence in it, and let resolved-market history tell you how often the pattern actually paid.
Compare the coins
The crypto data page covers all four assets and timeframes; the historical guide shows how to pull aligned series.
Frequently asked questions
Which crypto prediction markets can I correlate?
BTC, ETH, SOL, and XRP Up/Down markets, across every timeframe (5m, 15m, 1h, 1d). All four are captured on the same pipeline with millisecond timestamps, so their implied-odds series align cleanly for correlation work.
Why does timestamp alignment matter for correlation?
Correlation on mismatched timestamps produces misleading numbers. Because the four coins share one capture pipeline and clock, their series line up exactly, so the rolling correlation reflects real co-movement rather than a resampling artefact.
What can I do when correlation breaks down?
A breakdown often signals a coin-specific catalyst. Traders use it for relative-value ideas, fading the divergence if they expect the usual link to reassert, or to rotate toward whichever market is leading a co-ordinated move.
Does correlation differ across timeframes?
Yes, and meaningfully. Each coin is captured at 5m, 15m, 1h, and 1d, and the same pair can look tightly coupled on the daily series while diverging intraday. Compute the correlation on the timeframe your strategy actually trades, a 5-minute regime is a different thing from a daily one, and mixing them blurs both.