Intermarket Relationships

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. Use the ladder rung to set your trading stance. Assign one of these rungs based on what you found: - Rung 1: Coincidence - timing doesn’t lead reliably, or the move comes and goes with no persistence. Treat it as background noise. - Rung 2: Shared driver - both markets respond to the same event, and the “source” lead looks inconsistent. Trade only with the driver in mind, or trade the cleaner instrument directly. - Rung 3: Directional influence - the source leads, it persists, and you can state a mechanism. Trade the target with a plan for the lag. - Rung 4: Feedback or reflexive loop - the target’s move feeds back into the source (common in rates/credit during stress). You need extra confirmation; otherwise you’ll chase the loop.

The ladder doesn’t “prove” causality like a lab test. It forces you to stop treating correlation as destiny. When you can’t reach Rung 3, you reduce position size or you switch to a different signal pair.

Putting it into practice: Daria’s cross-market read with a causality trap

Daria trades a swing window of roughly one to three weeks. She’s been seeing a recurring pattern: when Treasury yields rise, she feels pressure to fade the rally in the S&P 500. She’s not wrong often - but she gets chopped when the relationship breaks. Last week, she watched yields firm early in the session, but the index didn’t follow immediately. She wanted to know: is this influence with a lag, or just correlation from the same news?

Here’s how she runs the Influence Ladder Model on her charts, using the same structure every time.

1. Choose the source and target clearly. Source: the 10-year Treasury yield (she tracks it daily). Target: the S&P 500 index (daily close). Assumption she writes down: “If yields reprice first for a mechanism-driven reason, the index should weaken or lag.”

2. Mark 6-10 past turning points, not the last one. She scrolls back and finds the last 8 occasions where yields made a clear swing in either direction. For each occasion, she marks: - the date yields turned - the date the S&P 500 turned meaningfully after Expected outcome: yields turn first often enough to matter.

3. Measure lead/lag with a simple date difference. She records the lag in trading days between the yield turn and the index turn. If the “lead” is usually within 0-3 days and the index then holds its response for 5+ trading days, she can consider Rung 3. If yields turn after the index or the lag looks random, she drops the confidence.

4. Check persistence: does the yield move hold? For the same occasions, she checks whether yields stay in the new direction for at least a week of sessions. If yields spike intraday and then snap back, she marks the event as likely coincidence.

5. Write the mechanism in one sentence. She doesn’t overcomplicate it. She writes: “Rising yields lift discount rates and tighten financial conditions, which pressures equity valuation.” If a specific event drove the yield move (like a sudden risk-off bid for safe havens), she adjusts the mechanism sentence: “Safe-haven demand can rise even when growth expectations fall, which hits equities through risk appetite.”

6. Decide what to trade based on the rung. - If she lands on Rung 2 (shared driver), she waits for the driver to confirm and she trades the instrument that reacts cleaner. - If she lands on Rung 3 (directional influence), she plans a staggered entry: she starts with a smaller position during the yield lead, then adds only after the index shows follow-through.

7. Set an execution trigger tied to the ladder, not to hope. She uses a simple trigger: she requires the S&P 500 to break the most recent minor swing low after yields hold their direction for 3-5 sessions. Expected outcome: she avoids shorting the index at the exact moment yields spike if the market still needs time to reprice.

Quick checklist (Daria uses it before every intermarket trade): - Identify source (lead candidate) and target (lag candidate) - Mark 6-10 past turning points and record lead/lag dates - Verify persistence: source holds direction for 5+ sessions - Write a one-sentence mechanism without hand-waving - Assign a ladder rung and size the trade accordingly - Use a trigger that confirms the lag market’s repricing

This time, she finds that yields turned first in 6 of 8 events, and the average lag clustered around a few sessions rather than random dates. She also sees persistence: yields held for about a week in most cases. She lands on Rung 3 most of the time. She trades the index with a lag confirmation trigger instead of forcing an immediate move. The chop fades because she stops treating “yields up today” as “index down tomorrow no matter what.”

Common causality traps (and how to fix them fast)

Shared-driver mirage You connect asset A to asset B, but a third factor drives both. Your chart looks like a clean line until the third factor changes. Do this: Build a “driver note” for each event you use in your ladder testing. When you see yields and equities move together, write the likely driver in one phrase (inflation expectations, growth fears, risk-off liquidity, central bank tone). Then check whether that driver stayed stable across the events that supported your influence rung. Not this: Trade the pair on the relationship alone when the event type changes. If yields rise because inflation expectations climb versus yields rising because risk-off bids hit safe havens, you don’t get to use the same causal story.

Reflexive feedback loop Sometimes the target market doesn’t just respond; it feeds back into the source. In stress, equity selling can worsen credit conditions and push yields, or rate moves can tighten credit, which then hits equities. Your “lead” becomes a loop and your timing expectations break. Do this: Label any relationship where the target’s major move coincides with immediate changes in the source as “possible feedback” and require extra confirmation. For swing trading, demand persistence in both directions: the source holds after the target moves, not just the first impulse. Not this: Add to a position every time you see the first tick in the supposed lead market. In a loop, the first tick often becomes noise.

Non-stationary relationships Intermarket influence can change as market regimes shift: volatility changes, liquidity changes, and policy expectations change. A relationship that looked like Rung 3 for a quarter can drop to Rung 2 the next quarter. Do this: Re-test your ladder on a rolling window. If your timing and persistence checks degrade (for example, fewer than half of recent events show consistent lead), you downgrade the rung and tighten your trade rules. Not this: Keep the same signal logic indefinitely because it worked “often.” Influence depends on conditions; correlation doesn’t.

The takeaway is simple and actionable: you can’t outsource causality to a chart overlay. You earn it with lead/lag evidence, persistence, and a mechanism you can state in plain terms. When you treat correlation as a hypothesis and influence as a ladder rung you must climb, your cross-market signals stop feeling like luck - and start feeling like a process you can run under pressure.

End of chapter one. 7 more chapters in the full book.