Intermarket Analysis for Futures Traders: Using Bonds, Dollar, and Commodities Together

Every futures trading blog has the same three intermarket charts: a falling dollar next to rising gold, rising yields next to falling stocks, and a copper/gold ratio overlaid on the 10-year yield looking suspiciously aligned. What those charts never show you is the stretch where the relationship inverted for eighteen months and anyone trading it mechanically got run over. Intermarket analysis is useful — it's also the single most misapplied concept in retail technical analysis, because the correlations it's built on are regime-dependent, not physical laws.
The Textbook Relationships (And Why They're Half the Story)
John Murphy's framework from the 1990s is still the reference point across four major asset classes — bonds, commodities, currencies, and equities — and it's still worth knowing because it's a reasonable macroeconomic context, not because it's a signal generator on its own. The traditional sequence: dollar weakness supports commodity strength (most commodities are dollar-denominated, so a cheaper dollar makes oil, gold, and copper cheaper for foreign buyers, pushing demand and price up), commodity strength signals inflation and eventually pressures bond weakness (rising yields), and bond weakness feeds into stock weakness as higher discount rates compress valuations and raise corporate borrowing costs. Gold moves inversely to real rates because it has no yield of its own, so every basis point the 10-year moves is an opportunity cost on holding bullion. The framework is also sometimes used as an early warning signal for sector rotation — money moving out of rate-sensitive sectors and into commodity-linked ones tends to show up in these cross-asset relationships before it's obvious in any single market's price action.
All of this is directionally true often enough to be worth knowing. The dollar/gold relationship is probably the most reliable of the group: when the DXY has strong directional trends, the rolling correlation with gold has sat in the -0.6 to -0.9 range for extended stretches. But "often enough" isn't "always," and the gap between those two is where traders lose money trusting a textbook chart overlay instead of checking the actual rolling correlation on the pair they're trading right now.

The 2022 Stock-Bond Correlation Flip Nobody Saw Coming (In Size)
The clearest recent proof that these relationships are regime-dependent: the 60/40 stock-bond correlation, which had been reliably negative for roughly 20 years (1999-2021), flipped sharply positive in 2022 and pushed to the highest reading in 50 years. Stocks and bonds fell together because the dominant macro driver changed — it was no longer growth shocks (where bonds rally as a hedge when stocks fall), it was an inflation shock, where both equity valuations and bond prices get hit by the same rising-rate move. A 60/40 portfolio built on the assumption that bonds hedge stock drawdowns had its worst calendar year since the 1970s.
By late 2025 and into 2026, the rolling 24-month correlation between S&P 500 returns and 10-year Treasury total returns was running in the +0.10 to +0.30 range — no longer deeply negative, not sharply positive either. That's the key mechanical point: the sign of the stock-bond relationship depends on whether growth or inflation is the dominant macro force at that moment, and that can flip inside a single year. Anyone running a strategy that assumes a fixed-sign relationship between equity futures and bond futures needs to be checking which regime they're actually in, not assuming 2015's relationship still applies in 2026.
Copper/Gold and the 10-Year: A Relationship That Broke, Publicly, in 2023
The copper-to-gold ratio is a commonly cited proxy for the 10-year Treasury yield — copper rising relative to gold signals growth optimism (industrial demand up, safe-haven demand down), which historically tracked rising long yields. In 2023 that relationship visibly broke: the copper/gold ratio fell sharply, leading a lot of macro traders to position for lower 10-year yields. Instead, the 10-year surged to multi-decade highs by October 2023. The ratio pointed one way; the actual bond market went the other, for months, loudly.
The lesson from that specific episode isn't "the copper/gold ratio is useless" — it's that any two-asset proxy for a third asset is a correlation of convenience, not a causal link, and it can decouple for a year or more when something else (in that case, persistent term-premium repricing and fiscal supply concerns) is driving the yield independently of the growth/safety signal copper and gold are picking up. If you're using any ratio trade as a signal rather than a confirming data point, you're one regime shift away from a bad trade.
Dollar and Commodities: True More Often Than It's False, But Watch the Magnitude
The dollar-commodities inverse relationship holds up better than most intermarket pairs because the mechanism is direct: oil, gold, and most industrial metals are priced in dollars globally, so dollar strength mechanically makes them more expensive in other currencies, which dampens demand at the margin. This is less a "sentiment correlation" and more a pricing-mechanics correlation, which is part of why it's sturdier. But the magnitude of the relationship varies hugely by what's driving the dollar move. A dollar rally driven by Fed hawkishness (rate differentials) tends to hit gold harder than a dollar rally driven by global risk-off flows into US Treasuries, because in a risk-off dollar rally, gold is getting a competing safe-haven bid at the same time the dollar is rising against it. In 2025-2026, CME Group's research desk noted gold and Treasury yields went through a stretch where both markets were pricing divergent inflation narratives — gold rising on debasement/independence-of-the-Fed concerns while yields fell on rate-cut expectations — before the two converged sharply in late January 2026 when sticky inflation data hit both markets at once. That's a case where two "textbook inverse" relationships (gold vs. real rates) were both running, just on different timelines, until a single data print synced them.
Where Correlations Go When You Actually Need Them: To 1
The most important practical fact intermarket analysis rarely states plainly: in a genuine liquidity event, cross-asset correlations tend to converge toward 1 (everything sells off together) regardless of the "normal" relationship. March 2020 is the textbook case — gold, which is supposed to be the safe-haven hedge against equity drawdowns, sold off alongside stocks in the first days of the panic because funds were raising cash across every asset class to meet margin calls, not because the gold/equity fundamental relationship changed. The diversification benefit intermarket relationships are supposed to provide is most reliable in calm-to-moderate stress and least reliable exactly when you need it most — in a liquidity crunch. If your risk model assumes bond futures will offset equity futures losses in a crash, stress-test that assumption against 2020's first week and 2022's full year, not against a 10-year average correlation number.
What This Actually Means for Position Sizing and Signal Use
Practically, intermarket analysis is better used as a context filter than a standalone signal. If you're long crude oil futures and the dollar is breaking out to new highs on real rate-differential strength, that's a headwind worth weighing into your risk, not a mechanical "close the trade" signal. If you're running a risk-on futures book and the stock-bond correlation has been positive for the past six months, don't assume your bond position is still hedging your equity exposure the way it would have in 2015 — check the realized correlation on a 60-90 day rolling window before you size the pair as a hedge. The practical habit that separates traders who use intermarket analysis well from those who get burned by it: check the current rolling correlation on the specific pair and timeframe you're trading, don't assume the chart-book relationship from a trading book published in 1991 is still the regime you're in today.
The Bottom Line
Intermarket relationships are real, mechanically grounded in many cases (dollar/commodities especially), and worth knowing. They are not fixed-sign laws you can trade mechanically across any time period. Financial markets are interconnected enough that these relationships do provide useful macro context, and they can flag major market turning points before a single-market chart does — but the dollar-gold inverse relationship, the yield-equity relationship, and the copper/gold-yield proxy have all gone through multi-month to multi-year stretches where they inverted or decoupled entirely. In each case, traders who were treating the old relationship as gospel took losses on the "obvious" trade. Use intermarket context to size risk and flag divergences worth investigating — not as a mechanical entry signal on its own.