Intermarket Relationships

What do you do when two markets move together and your P&L tells you they shouldn’t? You’ve seen it: bonds sell off and stocks follow, the dollar strengthens and oil drops, or a commodity rally “pulls” equities - until it doesn’t. Correlation feels useful, but influence requires a different read. This chapter gives you a way to separate “they moved together” from “one move helped cause the other,” and it shows you how to avoid the common causality traps that turn intermarket signals into expensive guesses.

You’ll walk away able to take two (or three) cross-market signals - say, Treasury yields and equity sectors, or the dollar and commodities - and test whether the relationship makes trading sense. You’ll also learn how to structure your chart notes so you can spot when you’re just seeing timing overlap, shared news, or reflexive feedback loops. The result: fewer “looks right” trades and more signals you can defend when price action gets noisy.

Why correlation tricks you, and what “influence” actually means

Most intermarket “connections” start life as correlation: two series rise and fall around the same time. The problem is that correlation has multiple parents. Shared news can move both assets without either one influencing the other. Market stress can push several instruments at once through risk appetite and liquidity. And some relationships invert after the first shock because traders reprice the whole chain.

Influence means something stricter: changes in one market’s condition systematically change the odds of another market’s future move, even after you account for the common drivers. Influence can still run through multiple steps, but it has a direction and a mechanism. For example, when Treasury yields rise because inflation expectations jump, discount rates and funding costs change for equity and credit. That chain can affect equity valuation and corporate behavior. Correlation would say “yields and stocks moved together.” Influence asks “why did the bond repricing change the equity payoff map?”

That distinction matters because traders pay different prices for different errors. If you trade correlation as if it’s influence, you’ll often buy or sell right when the shared driver flips. If you trade influence as if it’s correlation, you’ll miss early moves where the lead market has already repriced the mechanism while the lag market still looks “quiet.”

To make this concrete, consider a retail swing trader like Daria, 34. She watches a simple set: the S&P 500 index, the U.S. Dollar Index (DXY), and WTI crude oil. She notices that when DXY rises, WTI often falls. If she treats that as influence, she might short oil every time the dollar ticks up. But sometimes the dollar strengthens because traders flee risk, and oil drops because risk-off cuts demand expectations. In that case, DXY didn’t “cause” oil’s drop; both markets responded to the same risk shock. If she adjusts her read, she can still trade the relationship - just not by forcing a causal story where none exists.

The Influence Ladder Model gives you a practical way to sort these outcomes without pretending you can know the one true cause. You’ll assign each relationship a rung based on what must be true for influence to exist, then you’ll use that rung to decide how tightly you can trade the signal.

The Influence Ladder Model: a step-by-step way to test causality across markets

The Influence Ladder Model sorts relationships into rungs from “likely shared driver” to “directional mechanism.” You don’t need a PhD or a black box. You need a repeatable checklist you apply every time you connect two assets.

Use this approach on any pair: equities vs bonds, FX vs commodities, rates vs credit, and so on. The ladder uses three ingredients you can verify from your charts and your event calendar: timing, persistence, and mechanism.

1. Start with timing: who moves first, and by how much?

Check whether the “source” market consistently turns before the “target” market. Use a simple measurement: mark the swing low/high dates on both charts for the last 6-10 comparable events. If the supposed source often turns after the target, you don’t have influence - you have a coincidence problem.

2. Confirm persistence: does the source keep its direction?

Influence usually shows up as a sustained repricing, not a one-day blip. Track whether the source market holds its move for several sessions (for swing trading, think 5-15 trading days) while the target lags. If the source reverses quickly and the target still trends, you’re probably looking at shared news or a third driver.

3. Demand a mechanism: can you name the path from source to target in plain terms?

You must be able to write one sentence that links them without hand-waving. Example: “Higher yields raise discount rates, which pressures equity valuation.” If you can’t state a mechanism, you don’t have influence; you have a chart pattern.

4....