# DXY Correlation Regimes

A Breadth, Sentiment & External Data concept (Macro/intermarket) in the LuxAlgo Library, with 1 indicator implementation.

## What are DXY Correlation Regimes?

DXY correlation regimes describe how the relationship between the U.S. Dollar Index and other markets shifts over time. DXY is a fixed-weight basket of the dollar against six currencies, dominated by the euro at roughly 57.6 percent of the weight. Risk assets, gold, and commodities often trade inversely to the dollar, but the strength and even the sign of that [correlation](https://www.luxalgo.com/library/concept/correlation/) drift across months: sometimes stocks and the dollar fall together, sometimes dollar strength is exactly what caps a rally. A regime is a stretch where the relationship is stable enough to lean on, usually measured with a rolling correlation of returns.

Traders use the active regime as a filter or a divergence source rather than a standalone signal. Because the euro dominates the basket, DXY behaves close to an inverted EURUSD, which makes small failures of that mirror informative: when one makes a fresh extreme and the other fails to confirm, dollar-based divergence traders read cracks in the move. Rolling correlations lag by construction and regimes end without notice, so the read needs periodic re-verification.

The regime shifts have drivers worth naming. The popular dollar-smile framework describes the dollar bid at both extremes, global crisis (safe-haven flows) and strong US outperformance (rate differentials), and soft in the middle, which is one reason the equity-dollar correlation flips sign across macro eras. The classic couplings each have their own logic: gold's inverse relationship runs through real rates and the dollar's pricing role, commodity and emerging-market inverses through dollar funding, and Bitcoin's inverse coupling only emerged clearly in the macro-driven era after 2020, having been largely absent before, a reminder that every coupling is conditional rather than structural.

In practice the regime work is measurement plus humility. Rolling windows from one to six months form a ladder, with agreement across windows defining a usable regime; sign stability matters more than magnitude; and the euro's dominance means every DXY read deserves a cross-check against EURUSD itself, plus a [currency strength meter](https://www.luxalgo.com/library/concept/currency-strength-meter/) when the question is which legs are actually driving. Divergence tooling like the SMT DXY study automates the failed-mirror detection, and the broader dashboard belongs to [intermarket analysis](https://www.luxalgo.com/library/concept/intermarket-analysis/): knowing which assets are currently coupled tells you which charts must be consulted before a trade.

## How to identify the active DXY regime

A regime is a measured, persistent relationship; identification is the measuring and the persistence test.

1. Choose the asset pair to test, DXY against the instrument you trade, and work in returns rather than price levels.
2. Compute rolling correlations across a ladder of windows, commonly one, three and six months, so one window's noise cannot masquerade as a regime.
3. Classify the state: persistently negative across windows is an inverse regime, persistently positive a coupled one, and sign-flipping means no usable regime exists.
4. Cross-check the mirror: because the euro dominates the basket, verify DXY moves against EURUSD, and treat basket-versus-euro disagreements as breadth information about the dollar move.
5. Re-verify on schedule and after macro shifts: rolling correlations lag by construction, and the regime that justified a filter can be gone weeks before the window admits it.

## How it's calculated

Classifies the rolling correlation between an asset's returns and US Dollar Index returns into positive, neutral, or inverse regimes.

```
x_t = P_t / P_(t-1) - 1
y_t = D_t / D_(t-1) - 1
rho_t = ( Σ_(i=t-n+1..t) (x_i - x_bar) × (y_i - y_bar) ) / sqrt( ( Σ_(i=t-n+1..t) (x_i - x_bar)^2 ) × ( Σ_(i=t-n+1..t) (y_i - y_bar)^2 ) )
Regime_t = inverse if rho_t <= -theta, neutral if -theta < rho_t < theta, positive if rho_t >= theta

  P_t: asset closing price at bar t
  D_t: DXY (US Dollar Index) closing level at bar t
  x_t: asset return at bar t
  y_t: DXY return at bar t
  x_bar: mean of x_i over the window
  y_bar: mean of y_i over the window
  rho_t: rolling Pearson correlation coefficient at bar t (-1 to +1)
  Regime_t: correlation regime label at bar t
  n: rolling window length in bars (commonly 20)
  i: bar index inside the window
  theta: correlation threshold separating regimes (commonly 0.3)
  t: bar index
```

The rolling Pearson correlation is the standard core; log returns ln(P_t / P_(t-1)) are an equally common input and give near-identical results.

Window and threshold are conventions rather than published standards; windows of 20 to 90 bars and thresholds of 0.3 to 0.5 are typical.

Correlating price levels instead of returns overstates the relationship and is best avoided.

## How traders use it

- As a directional filter: in a strongly inverse regime, a rising dollar argues against fresh longs in euro pairs, gold, or other dollar-sensitive assets until the correlation itself weakens.
- As a divergence engine: comparing DXY against euro-heavy pairs for failed mirror moves, the same logic formalized in [Smart Money Technique divergence](https://www.luxalgo.com/library/concept/smart-money-technique-divergence/).
- As part of a broader [intermarket analysis](https://www.luxalgo.com/library/concept/intermarket-analysis/) dashboard: knowing which assets are currently coupled to the dollar tells you which charts to consult before taking a trade.
- Decomposed by currency: when DXY moves, a strength meter shows whether the euro leg or the smaller components drove it, separating broad dollar moves from single-pair events wearing the index's clothes.
- Paired with risk gauges: dollar bid alongside rising [VIX](https://www.luxalgo.com/library/concept/vix/) is the classic risk-off coupling, and the pair confirming each other distinguishes a macro flight from a rate-differential grind that touches risk assets differently.

## DXY regimes vs related cross-market tools

- **Correlation** (https://www.luxalgo.com/library/concept/correlation/): Correlation is the statistic; the regime framework is its disciplined application, returns not levels, windows in ladders, persistence requirements, and the standing assumption that every relationship is temporary. The statistic misleads exactly when used without the discipline.
- **Intermarket Analysis** (https://www.luxalgo.com/library/concept/intermarket-analysis/): Intermarket analysis is the whole discipline of reading linked markets, bonds, commodities, currencies, equities. DXY regime work is its dollar-specific slice, narrow enough to operationalize as a daily filter and divergence source rather than a research project.
- **Currency Strength Meter** (https://www.luxalgo.com/library/concept/currency-strength-meter/): The meter decomposes strength per currency across many pairs; DXY compresses the dollar into one fixed-weight number dominated by the euro. The meter answers which currencies moved; DXY answers how the headline dollar traded. Regime work benefits from consulting both.

## FAQ

### Why does EURUSD mirror DXY so closely?

Arithmetic, mostly. The euro carries about 57.6 percent of the index weight, so DXY behaves largely like an inverted EURUSD plus five smaller currency components. The mirror is expected; the information is in its failures. When EURUSD makes a new low but DXY cannot print a matching new high, the dollar move lacks breadth across the other basket currencies.

### How do you measure a DXY correlation regime?

Compute rolling correlation on returns, not price levels, over a chosen window, commonly a few weeks to a few months, and watch for stability of sign and magnitude. A reading that flips sign frequently means there is no usable regime. Because rolling windows lag, many traders require the relationship to persist across several windows before treating it as active.

### What is the dollar smile?

A widely used framework holding that the dollar strengthens at both macro extremes, in global crisis, when safe-haven flows bid it, and in strong US outperformance, when rate differentials do, while weakening in the benign middle. Its relevance here is that it explains why correlations flip: the same rising dollar means opposite things for risk assets depending on which side of the smile is active, so the regime label needs the macro driver attached.

### Does Bitcoin follow DXY?

Episodically. Before 2020 the relationship was weak and unstable; in the macro-driven era since, a pronounced inverse coupling has appeared for long stretches, Bitcoin trading like a high-beta risk asset against dollar liquidity, punctuated by months where the link loosens entirely. The honest procedure is the same as for any asset: measure the rolling relationship, require persistence, and never assume last year's coupling survived.

### Why do stocks and the dollar sometimes fall together?

Because the coupling depends on the driver. When easier policy expectations weaken the dollar and lift risk appetite, the inverse regime holds; when a genuine crisis bids the dollar as a haven while equities fall, the same inverse sign appears for opposite reasons; and when US assets lead global flows, dollar and equities can rise or fall together. The correlation number cannot distinguish these; the macro narrative attached to the regime does.

### What window length should the rolling correlation use?

A ladder rather than a choice: roughly 20 sessions for the fast read, 60 for the working regime, 120 for the backdrop. Short windows react quickly and lie often; long ones are stable and stale. Requiring sign agreement across the ladder before treating a regime as active filters most of the false shifts, at the acknowledged cost of arriving late to genuine ones.

## Implementations in the Library

- DXY Correlation Regimes (LuxAlgo): https://www.luxalgo.com/library/indicator/dxy-correlation-regimes/

## Related concepts

- Intermarket Analysis: https://www.luxalgo.com/library/concept/intermarket-analysis/
- Yield Curve: https://www.luxalgo.com/library/concept/yield-curve/
- Credit Spreads: https://www.luxalgo.com/library/concept/credit-spreads/
- Real Yields: https://www.luxalgo.com/library/concept/real-yields/
- Macro Event Playbooks: https://www.luxalgo.com/library/concept/macro-event-playbooks/
- ES Fair-value Basis: https://www.luxalgo.com/library/concept/es-fair-value-basis/

---

Source: https://www.luxalgo.com/library/concept/dxy-correlation-regimes/ (LuxAlgo Library, the encyclopedia of trading & technical analysis). Free to use with attribution: https://www.luxalgo.com/library/license/