Platform

Correlations, Hedging & Statistical Arbitrage

This page answers three questions, in escalating order of commitment. How do my markets relate to each other? That's the correlation heatmap. How do I offset an exposure I can't trade directly? That's the Hedge Pair Explorer. And when is a statistically linked pair temporarily out of line and likely to snap back? That's the StatArb Cointegration Matrix. Read the sections in that order — each one narrows a genuine relationship between two markets into something progressively more actionable, from a picture of concentration risk to a position you can actually size.

The correlation heatmap: 1Y vs 5Y

Strong blue cells mean two commodities tend to move together; strong red means they tend to move in opposite directions; pale cells mean little relationship either way, and the diagonal is always a perfect +1, since every commodity matches itself exactly. A period selector switches the whole heatmap between 1Y, how relationships look right now, and 5Y, the durable structural picture behind them. Read across any row to see how one commodity relates to everything else you follow, and hunt for two shapes in particular: deep-blue clusters, and red pairs sitting on their own.

If everything you buy or hold glows blue on this map, you're carrying concentration risk — one shock hits the whole book at once, and a bad month is bad everywhere simultaneously. Pale and red names spread that risk out instead, and red pairs in particular are your natural hedge candidates: when one side falls, the other tends to rise and cushion the blow. Comparing the 1Y view against the 5Y view answers a quieter, equally important question — whether a hedge you've been relying on still holds today, or has decoupled without anyone noticing. A pair tightly linked over five years but drifting apart this year, the kind of relationship copper and natural gas illustrate on the copper and natural gas pages, is a story worth investigating before you lean on it again.

Hedge Pair Explorer: from statistics to a verdict

The Hedge Pair Explorer is a table of commodity pairs that have moved together over the long run, and for a buyer it answers a specific problem: how do you cap the downside on a purchase you can't lock in directly, because there's no convenient contract or liquid instrument for that exact commodity? Each row carries the pair's correlation, a Cointegrated Yes/No flag, an ADF p-value, a Hedge Ratio, and a plain-English Recommendation. Filter to "Cointegrated: Yes" and "Strong hedge" to surface only the pairs dependable enough to build a position on. When you're exposed to a thinly traded commodity with nothing to hedge it directly, a cointegrated pair lets you offset it with a more liquid proxy that has historically tracked it closely.

The Hedge Ratio is a recipe, nothing more exotic than that: for every 100 units of commodity A, trade this many units of commodity B to offset the price risk. A Cointegrated flag of Yes means the two prices share a genuine, stable long-run link rather than a coincidence that will quietly drift apart, and the ADF p-value is the strength gauge behind that flag — lower is more reliable, and the platform does the comparison for you. None of this is investment advice, and none of it is Market Lens placing a trade: the platform does not route orders or execute anything on your behalf. What you get here is decision support — a hedge ratio you hand to whoever executes financial positions for your organisation as a specification, or a blueprint you validate yourself against your own book, liquidity and limits before you act on it.

StatArb Cointegration Matrix: trading the spread

The StatArb Cointegration Matrix takes the same hedge pairs and upgrades them into a trading screen for market-neutral mean reversion. When a statistically tethered spread stretches unusually wide, the trade is to buy the cheap leg and sell the expensive one, profiting from the gap closing rather than from which way the broader market moves. A window selector switches the lookback between 1Y and 5Y, the same choice as on the heatmap, so you can judge whether the tether you're trading is a current phenomenon or a durable one.

1. Scan the Z-score column — Look for the rows the platform has colour-flagged — those are the spreads that have stretched furthest from their normal range, and the ones most worth a closer look this session.
2. Follow the Signal — It names the cheap leg to buy and the expensive leg to sell, so there's no guesswork on direction once a pair earns your attention.
3. Match the half-life to your horizon — A short half-life points to a trade measured in days; a long one means being prepared to hold for weeks. Size your patience to the number, not the other way around.
4. Balance the legs with the hedge beta — The beta is the ratio between the two legs — how many units of one to hold against the other so the position is genuinely market-neutral rather than a disguised directional bet.
5. Export to War Room — Send the candidate into the scenario simulator and stress-test it under a macro shock before you commit any capital to it.

What that export actually checks is simple to state and easy to miss on your own: the simulator propagates a shock through the network of commodity correlations and shows whether both legs of the pair move together under stress — meaning the trade holds — or split apart — meaning the hedge breaks. That split is exactly the failure mode that matters most for a market-neutral position, and it's the one thing a static table of historical correlations can't show you by itself. See the War Room for the full walkthrough.