Research · Dataset
Gold in your currency, not in dollars
Almost every statement about gold is made in US dollars. If you are paid in euros, yuan, yen or pounds, your gold return is the dollar gold price plus what your currency did against the dollar. This page rebuilds gold in four local currencies from primary sources and measures how much of the result came from each leg.
Data through 2026-10 · Source data updated November 03, 2026
Source and method
Exchange rates are converted to local currency per US dollar
and the FRED unit conventions were checked against ECB reference rates
before use: DEXUSEU is USD per EUR, DEXCHUS is CNY
per USD, DEXJPUS is JPY per USD, DEXUSUK is USD
per GBP. Two of the four are inverted on the way in, and getting that
wrong would silently flip the sign of every result.
Prices are monthly averages, so a "local gold return" is the change in the monthly average price converted at the monthly average rate — not a price you could have transacted at. No transaction cost, financing, storage or local tax is included anywhere on this page.
The arithmetic, stated once
A local gold price is just the dollar price times the exchange rate, so in log terms the return splits exactly:
r(local gold) = r(dollar gold) + r(exchange rate)
This is not an approximation. The script recomputes it for every month of every currency and reports the largest residual; it is 0.000000000000 — floating point noise, not a modelling choice. Because the split is exact, the two legs can be attributed without ambiguity: whatever the currency contributed is what is left over.
What each currency actually got
| Currency | Sample | Local gold | Dollar gold | FX | Gold's share | FX's share |
|---|---|---|---|---|---|---|
| € Euro | 1999-01 → 2026-10 | +1,380.6% | +1,330.3% | +3.5% | 98.7% | 1.3% |
| ¥ Chinese yuan | 1981-01 → 2026-10 | +3,074.6% | +637.0% | +330.8% | 57.8% | 42.2% |
| ¥ Japanese yen | 1971-01 → 2026-10 | +4,666.6% | +10,702.6% | -55.9% | 121.2% | -21.2% |
| £ British pound | 1971-01 → 2026-10 | +19,450.4% | +10,702.6% | +81.0% | 88.8% | 11.2% |
The two right-hand columns are shares of the log return, so they add to 100% and can exceed it on one side — when the currency moved against you, gold had to work harder to deliver the same local result.
Read the yen row first. A Japanese holder's gold return over 55.75 years was +4,666.6% while the dollar return was +10,702.6% — the yen lost 55.9% against the dollar, so 21.2% of the dollar return was handed back to the currency. Gold was 121.2% of the story.
The yuan is the mirror image. A Chinese holder's +3,074.6% over 45.75 years came from a dollar gold return of only +637.0% plus a currency that gained 330.8% against the dollar — the currency contributed 42.2% of the local return, more than the gold itself.
The euro and sterling are the least dramatic: 98.7% and 88.8% respectively. Over those windows the currency did very little either way, and the dollar gold price was almost the whole story.
Across the four currencies the dollar gold price accounts for between 57.8% and 121.2% of the local-currency log return. Same metal, same month, same ounce — and the answer to "how did my gold do" can still differ by a factor of two depending on who is asking.
How often did local gold beat dollar gold?
Over the full monthly sample, in what share of months did local-currency gold return more than the dollar price? If the currency were irrelevant, this would be 50%.
The answer is a coin flip for every currency: 50.3% for the euro, 45.9% for the yuan, 51.3% for the yen, 49.6% for sterling. Months where the exchange rate did not move are excluded from the count.
The monthly correlation between dollar gold and local gold tells the same story from the other side: 0.831 for the euro, 0.860 for the yuan, 0.853 for the yen and 0.883 for sterling. High, but clearly below 1.0 — which is exactly what the decomposition predicts, because the currency adds an independent component to the monthly move.
Decade by decade
The full sample hides the fact that the currency leg flips sign repeatedly. Here is the same data split into periods.
| Currency | Period | Dollar gold | FX | Local gold |
|---|---|---|---|---|
| € Euro | 1999-1999 | -1.4% | +14.7% | +13.1% |
| € Euro | 2000-2009 | +299.7% | -30.5% | +177.7% |
| € Euro | 2010-2019 | +32.3% | +28.4% | +69.8% |
| € Euro | 2020-2026 | +163.0% | -0.9% | +160.6% |
| ¥ Chinese yuan | 1981-1989 | -26.6% | +169.5% | +97.9% |
| ¥ Chinese yuan | 1990-1999 | -31.0% | +74.9% | +20.7% |
| ¥ Chinese yuan | 2000-2009 | +299.7% | -17.5% | +229.6% |
| ¥ Chinese yuan | 2010-2019 | +32.3% | +2.7% | +35.9% |
| ¥ Chinese yuan | 2020-2026 | +163.0% | -3.4% | +154.1% |
| ¥ Japanese yen | 1971-1979 | +1,097.4% | -32.9% | +703.9% |
| ¥ Japanese yen | 1980-1989 | -39.4% | -39.6% | -63.4% |
| ¥ Japanese yen | 1990-1999 | -31.0% | -29.2% | -51.2% |
| ¥ Japanese yen | 2000-2009 | +299.7% | -14.6% | +241.4% |
| ¥ Japanese yen | 2010-2019 | +32.3% | +19.8% | +58.4% |
| ¥ Japanese yen | 2020-2026 | +163.0% | +44.6% | +280.2% |
| £ British pound | 1971-1979 | +1,097.4% | +9.3% | +1,209.0% |
| £ British pound | 1980-1989 | -39.4% | +41.8% | -14.1% |
| £ British pound | 1990-1999 | -31.0% | +2.4% | -29.4% |
| £ British pound | 2000-2009 | +299.7% | +1.1% | +304.0% |
| £ British pound | 2010-2019 | +32.3% | +23.3% | +63.1% |
| £ British pound | 2020-2026 | +163.0% | -1.6% | +158.7% |
The 1980s are the clearest illustration. Dollar gold fell -39.4% across the decade, but a sterling holder lost only -14.1% because the pound also fell against the dollar. A yen holder lost -63.4% — a worse outcome than the dollar holder, because the yen had already started appreciating in the late 1970s.
In the most recent period (2020 to 2026) the yuan and euro were close to flat against the dollar, so local gold tracked dollar gold almost exactly. The yen is the outlier again, up +44.6% against the dollar, which turned a +163.0% dollar gain into a +280.2% local gain.
Drawdowns in local currency
The deepest peak-to-trough fall in each currency's own gold price, against the same fall measured in dollars. This is where the currency leg matters most, because a drawdown is the moment a holder is most likely to sell.
| Currency | Local peak → trough | Local depth | Dollar peak → trough | Dollar depth |
|---|---|---|---|---|
| € Euro | 2012-09 → 2013-12 | -34.2% | 2011-09 → 2015-12 | -39.3% |
| ¥ Chinese yuan | 2011-09 → 2015-11 | -39.0% | 1981-01 → 1999-07 | -54.0% |
| ¥ Japanese yen | 1980-02 → 1999-09 | -82.6% | 1980-01 → 1999-07 | -62.1% |
| £ British pound | 1983-02 → 1999-08 | -50.0% | 1980-01 → 1999-07 | -62.1% |
The yen drawdown was -82.6% against -62.1% in dollars — a yen holder's gold fell far more than the dollar price did, because the yen was rising the whole way down. The euro's 2000s drawdown was -34.2% against -39.3% in dollars, which is the opposite case: the euro was falling, so it cushioned the dollar loss.
These are monthly averages, so intramonth extremes are not visible. A drawdown measured on daily closes would be deeper than every number on this page.
After inflation
Nominal local gold prices can only rise if the currency is not losing value. This section converts each currency's year-end gold price into a real price using that country's consumer price index, then indexes the result to 100 in the first available year.
| Currency | Sample | Real index (first year = 100) | Real change | Real max drawdown | Worst year |
|---|---|---|---|---|---|
| € Euro | 1999–2025 (27 yrs) | 759.7 | +659.7% | -31.2% | 2013 (-31.2%) |
| ¥ Chinese yuan | 1986–2025 (40 yrs) | 408.6 | +308.6% | -57.3% | 2013 (-31.1%) |
| ¥ Japanese yen | 1971–2025 (55 yrs) | 1,437.5 | +1,337.5% | -81.6% | 1975 (-31.2%) |
| £ British pound | 1971–2025 (55 yrs) | 1,261.8 | +1,161.8% | -69.9% | 2013 (-30.1%) |
These are the numbers that matter most for a holder who spends the money. A sterling holder's real gold gain since 1971 is +1,161.8%, and a yen holder's is +1,337.5% — even though both watched the same ounce of metal roughly double and a half in dollar terms.
The real drawdowns are severe in every currency: -81.6% for the yen, -69.9% for sterling, -57.3% for the yuan and -31.2% for the euro. A "gold as inflation hedge" claim has to survive these numbers, not the nominal chart.
CPI sources differ by currency: FRED CP0000EZ19M086NEST (Euro area HICP, monthly) for the euro area, and the World Bank annual consumer price index for the others. The euro-area series is annual here (December values) because the World Bank API returns null for the Euro area aggregate, so FRED is the only usable source for it.
Where the four currencies stand now
| Currency | Month | Dollar gold | FX (per USD) | Local gold | 1-month | 1-year |
|---|---|---|---|---|---|---|
| € Euro | 2026-10 | $4,105.00 | 0.8930 | € 3,666 | -2.4% | +5.2% |
| ¥ Chinese yuan | 2026-10 | $4,105.00 | 6.6844 | ¥ 27,439 | -5.3% | -5.0% |
| ¥ Japanese yen | 2026-10 | $4,105.00 | 157.9748 | ¥ 648,486 | -4.0% | +5.6% |
| £ British pound | 2026-10 | $4,105.00 | 0.7523 | £ 3,088 | -4.2% | +1.6% |
The one-year column is the live version of the page's argument: over the past twelve months dollar gold moved +1.2%, but a yuan holder saw -5.0% and a euro holder +5.2%. The metal did the same thing for everyone; the currency decided who got paid.
What this dataset supports, and what it does not
Supported: that the currency is a first-order part of a gold holding for anyone who does not think in dollars. Over the full samples the currency leg supplied between 1.3% and 42.2% of the local-currency log return; the yen holder's real drawdown was -82.6% against -62.1% in dollars; and monthly local gold beat monthly dollar gold roughly half the time in every currency, which is what an independent FX component looks like.
Not supported: that this makes gold a better or worse investment for any of them. This page measures monthly average prices with no carry, no financing, no storage, no local tax and no bid/ask, and it says nothing about a holder's own currency hedging. It also does not predict the next period — the decade table shows the FX leg flipping sign repeatedly, and no part of this data identifies which way it goes next.
The practical reading is narrower and still useful: quote gold and judge gold in the currency you actually spend. A dollar-quoted gold chart is a different asset from the one a yen holder owns.
Frequently asked questions
Does holding gold in a local currency beat holding it in dollars?
Sometimes, and the difference is almost entirely currency rather than gold. For the euro over 334 months from 1999-01 to 2026-10, local gold returned 10.2% annualised against 10.1% for dollar gold, and the currency contributed 0.1% annualised. In log-return terms 98.7% of the move was the gold price and 1.3% was the exchange rate. For the renminbi the split is far more balanced: 57.8% gold against 42.2% currency, and for the yen the currency contributed -21.2%, a drag on the dollar return.
How often does local-currency gold outperform dollar gold?
About half the time, which is the honest answer to a question that gets answered with "it depends on the currency". Over 333 euro months, local gold beat dollar gold in 50.3% of them. The renminbi figure is 45.9% and the yen 51.3%. The currencies cluster tightly around 50%, which is what you would expect if the currency contribution is close to a coin flip over monthly horizons.
How closely do dollar and local gold moves track each other?
Very closely. The monthly log-return correlation between dollar gold and euro gold is 0.8307 with a 95% interval from 0.7941 to 0.8613 over 333 months. The renminbi correlation is 0.8596, the yen 0.8534 and sterling 0.8827. So converting gold into a local currency does not meaningfully diversify it; it overlays a currency return on top of a gold return that is already highly correlated.
What is the current level of gold in each of these currencies?
In 2026-10, gold was 4,105 per ounce in dollars, 3665.94 euros, 27439.34 renminbi, 648486.40 yen and 3088.08 pounds. Over 12 months the euro price rose 5.2% against 1.2% for the dollar price, while the renminbi price fell -5%. That divergence is the whole point of the page: identical ounces, very different local-currency outcomes.
How large have local-currency gold drawdowns been?
Larger than the dollar series in every case, which is the cost of the currency overlay. Euro gold fell -34.2% from its 2012-09 peak to its 2013-12 trough, against -39.3% for dollar gold. The renminbi series fell -39% and the yen -82.6%. All of these recovered, but the local-currency investor sat through a deeper and longer trough.
What does gold look like after adjusting for local inflation?
Better in the long run for most of these currencies, with real drawdowns that are still severe. In real terms the euro series rose 659.7% since 1999 on FRED CP0000EZ19M086NEST (Euro area HICP, monthly), with a worst year of -31.2% in 2013 and a best year of 42.8% in 2025. The renminbi's worst real year was -31.1% and its real drawdown -57.3%; the yen's was -81.6%.
How is the local-currency gold price actually computed?
By the exact identity local gold = dollar gold multiplied by the exchange rate, with no fitting or smoothing. The check is run on every rebuild: across 334 months from 1999-01 to 2026-10 the maximum absolute error is 0. The identity is exact, which means every figure on this page is reproducible from two published series rather than from a model.
Sources
- World Bank Pink Sheet, "Gold ($/troy oz)", Monthly Prices sheet, column 70 — Monthly Prices sheet, column 70, updated November 03, 2026. https://www.worldbank.org/en/research/commodity-markets
-
Exchange rates: Federal Reserve H.10 release, retrieved via FRED
(
DEXUSEU,DEXCHUS,DEXJPUS,DEXUSUK). Unit conventions cross-checked against ECB reference rates. https://fred.stlouisfed.org/series/DEXUSEU - Consumer prices: Euro area HICP via FRED; annual CPI via World Bank Development Indicators FP.CPI.TOTL. https://data.worldbank.org/indicator/FP.CPI.TOTL
-
Analysis script:
scripts/gold_fx_local_currency_analysis.py. Independent recomputation from the published CSVs (181 checks, separate numpy implementation):scripts/verify_gold_fx.py. - For the dollar-side fundamentals that drive the common component of all four series, see Gold vs. real yields and the dollar, Inflation-adjusted gold and Gold price history.
Data through 2026-10 (annual real series through 2025). Prices are nominal unless stated. Nothing on this page is investment advice or a forecast.