Direct answer
When people ask what forex pairs correlate with gold, the most commonly tested pairs are those that involve the US dollar or other major currencies that tend to move with global risk sentiment, interest-rate expectations, and inflation expectations. In practice, analysts often examine EUR/USD, USD/JPY, and other major “high liquidity” pairs (such as GBP/USD or AUD/USD) against gold’s price changes. Correlation is not guaranteed and may be weak or unstable depending on the market environment and the time period.
How the correlation typically works
“Correlation” here means a statistical relationship between two series, usually measured using returns (percentage changes) of gold versus percentage changes of a forex pair over the same dates.
A key driver is the role of the US dollar in global markets. Gold is frequently discussed as a USD-linked asset in the sense that gold’s price often responds to USD moves, though the exact direction and strength can vary by regime. Because of that, pairs that move strongly with the USD are usually the first candidates for correlation testing.
Another set of factors is rates and macro expectations. If expectations for real yields, inflation, or growth shift, both gold and currency pairs may react—sometimes in the same direction, sometimes not. In addition, risk sentiment can affect both gold (often viewed as a hedge in some periods) and currencies (for example, safe-haven demand). These mechanisms explain why correlations are worth checking, but they also explain why they can change.
Example checks you can run (and what to compare)
To independently verify whether a forex pair correlates with gold, you can compare gold returns with pair returns over a consistent timeframe, using a method like rolling correlation.
Vergelijkcriteria (and both options per criterion):
- Time granularity: use either daily returns or intraday returns (pick one and keep it consistent).
- Direction: use either gold returns versus pair returns aligned by date, or examine both directions (because correlation is about association, not causation).
- Window length: use either a shorter window (to see recent behavior) or a longer window (to see broader tendencies).
What you should expect in many periods is not a single permanent ranking of “best correlated” pairs. Instead, you may see periods where USD-related major pairs show a stronger association than other majors, followed by periods where the relationship weakens.
If you specifically test EUR/USD and USD/JPY against gold, you may find that the relationship changes sign or magnitude across different regimes—this is a common outcome in correlation studies.
Limitations and uncertainty
- No fixed correlation: Correlation can be unstable across timeframes because macro drivers shift.
- Correlation is not causation: Even if gold and a pair move together, that does not prove a direct causal link.
- Data handling matters: Using different trading hours, holidays, or inconsistent return calculations can change results.
- Regime dependence: Major news events (rates expectations, inflation surprises, risk-off/risk-on swings) can temporarily dominate the relationship.
- No future inference: A correlation measured in the past does not imply a similar relationship in the future.
Given these limits, the most verifiable conclusion is about whether association is present in a specific tested period, not about a permanent “correlated with gold” label for any forex pair.