Does Gold Really Hedge Inflation?

The theory everyone starts with

The intuitive case for gold as an inflation hedge goes: paper currency loses purchasing power when prices rise, gold can't be printed, so gold should rise with inflation. It's a clean story, and it does show up in some multi-decade comparisons. The problem is the mechanism it leaves out. Gold pays no interest, no dividend, and no coupon. Holding it means giving up whatever a bond or savings account would have paid you. That giving-up-of-yield is called the "opportunity cost" of holding gold, and it isn't set by inflation alone — it's set by the real interest rate, meaning the nominal yield on a bond minus the inflation rate investors expect over its life. When real rates are very negative, holding non-yielding gold costs you little, or nothing. When real rates turn strongly positive, gold competes with an asset that pays you to wait, and that competition can pull money away from gold even while headline inflation stays high.

2022: inflation stayed high, gold fell anyway

2022 is the cleanest recent test case because inflation and real rates moved in opposite directions at the same time. The U.S. 10-year real yield (measured by the TIPS-based DFII10 series) went from deeply negative to solidly positive over the year, while gold's price fell.

Read the two columns together: as the real yield climbed from -0.91% to 1.69% over ten months, gold's price fell by roughly 15% from its April peak. This happened while inflation was not cooling — the CPI index we track went from 282.5 in January 2022 to 302.8 by April 2023, a continued climb. Inflation was elevated the entire time; what changed was that bonds suddenly offered a positive, inflation-adjusted return, so gold's cost of doing nothing went up sharply. That is the mechanism in isolation, visible in a single year.

The 1970s-80s: the original counterexample

The same pattern, at larger scale, is why "gold hedges inflation" became conventional wisdom in the first place, and also why economists have argued about it ever since. Our own historical CPI series shows just how fast prices were rising in that stretch: the index climbed from 73.0 in July 1979 to 94.4 by January 1982 — about 29% in two and a half years, one of the fastest sustained inflation periods in the post-war data. It's also the period most often cited for the opposite lesson: once the Federal Reserve under Paul Volcker pushed nominal interest rates decisively above the inflation rate, turning real rates strongly positive, gold's price fell for years even though inflation stayed elevated for a while longer. We don't have gold price data from that era in our own database — our GC_F series only extends back to the mid-2010s — so we won't put a number on that decline here. What we can say, from the CPI figures above, is that the inflation side of the story is well documented; the reason gold didn't simply track it is the same real-rate mechanism at work in 2022.

What today's data actually shows about the link

If real rates are the real driver, current numbers should show gold responding more to real yields than to inflation expectations. Our correlation data (through September 18, 2026) is consistent with that, with an important caveat about timing. Over 60-day rolling windows, gold's daily moves and the 10-year real yield's daily moves have a correlation of -0.52 — a fairly strong inverse relationship. But the same correlation measured day-to-day over 21 days is only -0.18, weak enough to look almost like no relationship at all. That gap isn't a contradiction: real yields shift slowly, so a same-day correlation is the wrong tool to see the effect, while a 60-day window captures the drift. Meanwhile, gold's correlation with 10-year inflation breakeven rates — the market's own forecast of future CPI, currently 2.33% — is essentially flat over the past year (0.02) and slightly negative over the past month (-0.22). In plain terms: right now, gold is tracking real rates more than it's tracking inflation expectations, which is the opposite of the simple hedge story.

It's also worth naming the limit of that reading. The current real 10-year yield, 2.68% as of September 16, 2026, sits at the 99.6th percentile of its own past year — historically high, in a regime our data labels "restrictive" — yet gold has not collapsed to its one-year low; it sits roughly a third of the way up its 12-month range. So the real-rate channel is active, but it's not the only force on the price; central bank demand, currency moves, and positioning all matter too, and none of them are captured by a single correlation number. Anyone building expectations for a specific inflation print or rate decision should check the mechanism against the calendar of upcoming releases at macrogold.ca/calendar rather than assume last cycle's pattern repeats automatically.

The takeaway, and what it doesn't tell you

Gold has sometimes moved with inflation and sometimes against it, and the difference in every documented case comes down to what real interest rates were doing, not what headline inflation was doing. That is a mechanism, not a forecasting rule: knowing that real rates drive gold doesn't tell you where real rates are headed next, and our own backtests have found no measurable, repeatable edge in trading gold off any single inflation or rate indicator. The data here also has real boundaries — our correlation windows cover recent years, not every historical regime, and our gold price history doesn't reach back to the 1970s-80s episode that shaped the popular narrative in the first place. Treat "gold hedges inflation" as a conditional statement — true when real rates are falling or deeply negative, unreliable when they're rising — rather than as a fixed property of the metal.