The relationship between the US dollar (usually proxied by the DXY trade-weighted index) and US equities is one of the most cited — and most genuinely unstable — links in intermarket analysis. The central tension is that the sign of the correlation flips with the macro regime: a strong dollar mechanically depresses the foreign earnings of S&P 500 multinationals (an inverse, fundamental channel), yet in panics the dollar and equities can move together in opposite directions (risk-off dollar strength + falling stocks) or, in some growth booms, together in the same direction (US outperformance lifts both). There is no single, durable "dollar up → stocks down" law, and treating one as if it exists is the most common mistake practitioners make with this pair.

The mechanisms (why the sign changes)

Three distinct channels operate simultaneously, and which one dominates determines the observed correlation:

1. Translation / competitiveness channel (inverse). A large minority of S&P 500 revenue is earned abroad — commonly cited figures range from ~28% (Goldman Sachs, 2024, narrower geographic-segment basis) to ~40%+ (FactSet, broader basis), with the spread driven by methodology rather than disagreement on the direction; Information Technology, Materials, and Energy run materially higher (FactSet has cited IT near 56% foreign revenue). A stronger dollar converts those foreign sales into fewer dollars at the translation line and makes US exports pricier abroad. This is the basis of the popular "strong dollar = headwind for earnings" view (StockCharts ChartSchool; Hartford Funds).

2. Safe-haven / risk-off channel (co-incident decline). The USD is the primary reserve currency. In crises, global investors liquidate risk assets and pile into Treasuries and dollar cash, so the dollar rises while equities fall — not because the strong dollar caused the fall, but because both respond to the same fear. This is the "dollar smile," a framework attributed to Morgan Stanley currency strategist Stephen Jen (c. 2001): the USD strengthens at the left (global panic) and right (US growth outperformance) of the smile, and weakens in the calm middle (dollar smile overview; Eurizon SLJ Capital).

3. Capital-flow / growth channel (positive co-movement). When the US economy and US rates outperform the rest of the world, capital flows into dollar assets, lifting both the dollar and US equities together. This positive association has been documented in some academic windows (e.g. Kim for 1992–2002, see below), though it is episodic rather than permanent.

How it's used in practice

• Earnings-revision lens. Sell-side analysts use a rough rule of thumb — figures vary materially by source and methodology, commonly cited in the range of a low-single-digit percent EPS hit per ~10% move in the trade-weighted dollar (in the opposite direction), with international-heavy sectors swinging more. Treat the exact coefficient as folklore, not a measured constant. The effect is also lagged, because many large caps hedge near-term cash flows, so a dollar move shows up in reported earnings only as hedges roll off.
• Sector/relative rotation. A persistently weak dollar is treated as a tailwind for multinationals, materials, and especially non-US equities — foreign developed and emerging-market ETFs show a strong negative correlation to the DXY, both from translation and from EM dollar-debt dynamics (Investing.com).
• Regime read, not a signal. Sophisticated users treat the dollar as one input into what regime we are in (risk-on growth, risk-off panic, US-led boom) rather than as a standalone buy/sell trigger for the S&P.

Adoption, debate & evidence

This is a mainstream concept — John Murphy ("the father of intermarket analysis") popularized the dollar's place in the four-market framework, and it is standard on StockCharts and across sell-side macro desks. But the evidence is genuinely contested on direction:

• The widely repeated technician claim is an inverse S&P–dollar relationship.
• Yet the academic record does not support a stable inverse sign. Bernard & Galati's BIS study (Aug 2000), reviewing two decades of data, concluded there was little evidence of a robust, significant correlation between US equity indices and the major exchange rates over long horizons — while noting that sharp contemporaneous co-movements occur in some episodes (e.g. heightened correlation in 1999). Separately, Kim, "The US stock market and the international value of the US dollar," documented a positive association over 1992–2002 (rising equities coinciding with a stronger dollar — the capital-flow channel). The takeaway: neither a clean inverse nor a clean positive sign holds across all periods.

The honest synthesis is that the correlation is regime-dependent and time-varying, not signed. The earnings/translation channel is real and measurable at the company level; the index-level directional correlation is not stable enough to trade mechanically. The dollar-smile framework itself has shown cracks: through 2008–2022 the USD reliably bid in risk-off, but in April 2025 (the tariff episode) the dollar and US assets fell together when the US was the source of instability — the safe-haven bid evaporated (dollar smile reassessment). Claims that a dollar bull cycle "delivers ~13% annual S&P returns vs ~8% in weak-dollar cycles" circulate widely but rest on small samples (a handful of decade-long cycles) and should be treated as folklore, not a robust base rate.

Strengths & limitations

Works best as a fundamental explanation of multinational earnings pressure and as a regime classifier when combined with rates, credit spreads, and breadth. The translation channel is the most defensible piece — it is an accounting fact, not a correlation artifact.

Fails when used as a fixed-sign mechanical signal. The single biggest misuse is asserting "dollar up, therefore stocks down" without first identifying the driver of the dollar move: a growth-driven rally and a panic-driven flight produce opposite equity outcomes despite both being "dollar up." A second pitfall is ignoring the lag — currency effects on earnings are deferred by hedging, so the market often reacts to forward EPS revisions, not the spot DXY.

Sources

• John J. Murphy, Intermarket Analysis: Profiting from Global Market Relationships (Wiley) — foundational framework.
• StockCharts ChartSchool — Intermarket Analysis
• Bernard & Galati, The co-movement of US stock markets and the dollar, BIS Quarterly Review, Aug 2000 — finds little evidence of a robust significant correlation over long horizons (refutes a stable inverse sign; not a clean positive finding either).
• Kim, The US stock market and the international value of the US dollar, J. Macroeconomics — positive association 1992–2002.
• Hartford Funds — Dollar Dynamics — earnings/translation channel; foreign-revenue exposure (figures vary: Goldman ~28% / FactSet ~40%+ by methodology; IT/Materials/Energy higher).
• Dollar smile — framework attributed to Stephen Jen (Morgan Stanley, c. 2001): overview, Eurizon SLJ Capital, and reassessment / April-2025 exception.

Disputes flagged: popular technical analysis asserts an inverse S&P–dollar relationship; the BIS (Bernard & Galati, 2000) finds no robust long-horizon correlation, while Kim finds a positive association in 1992–2002 — the truthful resolution is regime-dependence, not a fixed sign. The per-10%-dollar EPS "rule of thumb" and dollar-cycle return figures are widely cited but rest on rough/small-sample estimates and should not be treated as measured constants.