Hedging With Gold: What the Evidence Shows and Where It Fails

Short answer: gold is a partial hedge with four different jobs, and each one fails in its own way. Against the dollar it worked on most days in our data (gold rose on 71% of dollar-down days). Against inflation it worked over the very long run (9.0% a year against 3.9% CPI from 1970 to 2025) but lost to inflation through the 1980s and 1990s. Against stock crashes it was unreliable in the first days (up on 5 of the 10 worst S&P days in two years) and better over months. In a trading account, “hedging” a losing position with an opposite one freezes the loss and adds costs; the real hedge for a trade is a stop and the right size.

In plain words: a hedge is an umbrella. Gold is an umbrella that opens most of the time when the dollar rains, sometimes when inflation rains, and often a few minutes late when stocks pour. It’s still worth carrying, but you should know which weather it’s made for, and never assume it will open on time.

Where this lesson fits. It builds on two of our studies: is gold a good inflation hedge? and does gold go up when stocks crash?, and on the intermarket and correlation lessons (gold intermarket, correlation trading). This lesson puts the evidence together, adds our own counts, and draws the line between hedging a portfolio and hedging a trade.

What a hedge is, and what it is not

A hedge is a position you hold because it tends to gain when something else you own loses. A perfect hedge cancels the risk completely. Real hedges are imperfect and come in two strengths:

• A hedge in the strict sense reliably moves against the risk, in the same period, by a predictable amount. Futures used by a gold miner to lock in a selling price are an example.
• A diversifier tends to behave differently from your other assets, so the total swings less, but it can fall at the same time. Gold for a stock investor is mostly this.

Three questions decide whether gold works as a hedge in a given situation:

1. Against what? The dollar, inflation, a stock crash, a currency collapse, a banking crisis.
2. Over what time? A day, a month, a decade.
3. At what cost? Gold pays no interest, and gold positions on margin cost financing.

The rest of this lesson takes each risk in turn, with data.

Gold and the dollar

Gold is priced in US dollars. When the dollar weakens, gold becomes cheaper for buyers using euros, rupees or yuan, and its dollar price tends to rise. That is the most consistent of gold’s hedge relationships.

Our count over 501 trading days (4 November 2024 to 3 November 2026), using the ICE dollar index:

• Dollar up: gold fell on 160 of 259 days (62%), median gold change -0.30%.
• Dollar down: gold rose on 167 of 234 days (71%), median +0.61%.
• Big dollar days (0.5% or more): gold fell on 30 of 43 big up days (70%) and rose on 47 of 58 big down days (81%).
• The site’s correlation tool measured a 250-day correlation of -0.38 between gold’s and the dollar index’s daily returns.

So for someone whose wealth is in dollars and who worries about the dollar losing value against other currencies, gold has been a reasonable day-to-day offset. For someone whose wealth is in another currency, the picture is different: gold in rupees or yen also moves with the local currency. Our Tokyo session lesson shows gold in yen moving the other way from gold in dollars on some days.

Where it fails: on about one day in three, gold and the dollar moved in the same direction. When gold’s own drivers dominate (a central bank buying spree, a safe-haven rush, a real-yield shock), gold can rise with the dollar or fall with it. The gold intermarket lesson explains why real yields often matter more.

Gold and inflation

The claim: gold keeps its purchasing power when prices rise. The test: compare gold’s yearly change with US consumer prices (CPI, December to December) from the site’s history file, which goes back to 1968.

Two lessons jump out. Over the full 56 years, gold beat inflation comfortably. But an investor who bought gold at the end of the 1970s lost to inflation for twenty years. In the site’s inflation-adjusted series (in 2026-09 dollars), gold’s 1980 average price of about $2,094 was not beaten until 2010. The starting price decides most of the result.

The high-inflation years

If gold were a dependable inflation hedge, it should do best in the years when inflation is highest. It didn’t.

Of the 57 years from 1969 to 2025, 14 had US inflation of 5% or more. Gold beat inflation in 7 of them, almost all in the 1970s. In 2021 and 2022, with inflation at 7% and 6%, gold fell slightly both years. Our study on gold as an inflation hedge explains why: when a central bank fights inflation with fast rate rises, real yields rise, and gold, which pays no interest, loses appeal. The real yields and dollar tool tracks that driver live, and the gold and CPI tool shows each CPI release against gold.

Where it fails: over the 1 to 10-year horizons most people care about, and whenever inflation brings aggressive rate rises.

Gold and stock market crashes

The claim: when stocks crash, gold rises. The evidence: often over months, often not in the first days.

The worst stock days in our data

Out of 500 trading days from 4 November 2024 to 3 November 2026, the S&P 500 fell 2% or more on 10. Gold rose on 5 of them, with a median change of +0.19%. Gold rose on 54% of all days in the period, so on the worst stock days it behaved about as it does on any day. The site’s 250-day correlation between gold and the S&P 500 was 0.33: slightly positive, not negative. On 30 October 2026 the two fell together (-1.8% for the S&P 500, -3.3% for gold).

A real crash played out: April 2025

The early-April 2025 tariff shock is the clearest real example in our data of how gold behaves in a crash.

• 2 April 2025: S&P 500 5,670.97, gold (COMEX) 3,166.2.
• Phase 1, the dash for cash (to 8 April): stocks fell to 4,982.77 (-12.1%). Gold did not rise; it fell to 2,990.2 (-5.6%). Investors sold what they could, including gold, to raise cash and meet margin calls.
• Phase 2, the haven bid (to 21 April): gold rallied +14.6% from its low to 3,425.3, +8.2% above where it started, while the S&P 500 at 5,158.20 was still -9.0% below.

The same two-phase pattern appears in the bigger crises described in our crash study: on month-end closes, gold fell from 972.1 at the end of February 2008 to 716.8 at the end of October 2008, about -26%, before more than doubling by 2011; in March 2020 it fell about 12% in the ‘dash for cash’ before a record high in August. And in 2022, when stocks fell because the Fed was raising rates fast, gold fell too (1,795.0 at the end of January to 1,640.7 at the end of October).

Where it fails: in the first days of a liquidity panic, and in sell-offs caused by rising interest rates.

A small gold slice in a stock portfolio

What does that mean for someone holding stocks? We compared buy-and-hold mixes over the two years of data we have for the S&P 500.

The gold slice cut the worst drawdown by about two points for every 10% held, with little change in return, in this window. But this is one window of two years with a strong gold market, and gold itself fell 29.1% from its January 2026 high. The S&P figures exclude dividends. This is a description, not a recommendation; allocation is a personal decision.

Gold in crises and currency collapses

Gold’s strongest case as a hedge is in local currency crises and against policy risk. When a currency collapses, gold priced in that currency rises, because gold’s dollar price stays roughly where it was while the local currency falls. Central banks also buy gold to hold a reserve asset that is no other country’s liability; our guide to central bank gold buying covers it. Those uses are real, but they are about holding value over years, not trading.

Where it fails: gold is still a market. It can fall sharply in a crisis that triggers forced selling, and its price can drop more than 3% in a day, as on 30 October 2026. In a crisis you may also face wide spreads and trading limits.

Portfolio hedging vs trading hedges

The word “hedge” means two quite different things, and mixing them causes expensive mistakes.

A long-term holder who owns gold “as a hedge” is diversifying. A trader who opens a gold position is speculating, and the hedge for that speculation is not more gold.

Hedging a gold trade properly

For a XAUUSD trader the tools are, in order of simplicity:

1. A stop loss and correct size. The hedge that is always available, costs nothing to hold and caps the loss. The risk management lesson shows how to size from the stop.
2. Reducing size. If you are unsure, trade smaller. Half the size is half the risk.
3. Options. A put option protects a long position below a strike for a known premium. Our gold options guide explains the basics; options cost money and expire.
4. A correlated position. Long gold and short silver, for example, removes much of the shared metal risk and leaves the relative bet (see the gold/silver ratio lesson). It only works while the correlation holds.

Hedging inside your own account: why it usually backfires

Many platforms let you hold a long and a short XAUUSD position at the same time (“hedging mode”). It feels like protection. It isn’t.

Worked example (hypothetical): you are long 0.10 lot from 4,050.00. Price drops to 4,020.00, a $30 loss: 30 × 0.10 × 100 = $300. Instead of closing, you open a short 0.10 lot at 4,020.00.

• Your net exposure is now zero: every $1 move adds $10 on one leg and takes $10 from the other.
• The $300 loss is locked. It is exactly what you would have if you had closed the long.
• You paid a second spread to open the short, and will pay two more to close both.
• Overnight, you are charged financing on both legs. On most platforms both swaps cost money.
• The decision you were avoiding is still there: to get out, you must close one leg first, which leaves you with an open position in one direction. Most people close the “winning” leg, hold the loser and are back where they started.

That is why our test of grid and hedging systems ended badly: see gold grid and hedging risks for what happened to real-data accounts. If your platform uses a netting account, an opposite order simply closes the position; our MT5 hedging vs netting guide explains the difference.

Measuring a hedge: correlation, beta and hedge ratio

Professionals don’t ask “is gold a hedge?” but “how much does gold move when the thing I fear moves, and how reliably?” Two numbers answer it:

• Correlation (from -1 to +1) says how consistently two markets move together or apart.
• Beta says how far one moves, on average, for a 1% move in the other.

The site’s correlation tool publishes both for the last 90 trading days. To 4 November 2026, gold’s 90-day correlation with the S&P 500 was +0.43 with a beta of +0.82: in that stretch gold tended to move with stocks, by about 0.8% for each 1% move in the S&P 500, which is the opposite of a hedge. Against the dollar index the correlation was -0.44 and the beta -2.04: gold tended to fall about 2% for each 1% rise in the dollar index.

A hedge ratio turns this into a size: hedge ratio ≈ beta. If you wanted gold to offset a dollar exposure, the beta says how many dollars of gold offset one dollar of that exposure, on average, in the period measured. The catch is in the last four words. Betas and correlations drift, as the 90-day figures on the correlation tool show week by week, so any hedge ratio needs re-checking and a margin for error.

A quick hedge-ratio example (hypothetical): someone who spends euros holds $20,000 in dollar cash, which loses about 1% of its euro value for every 1% fall in the dollar. With a gold-dollar beta of -2, a 1% dollar fall has lifted gold about 2% on average, so roughly $10,000 of gold would offset the $200 loss. Check the sign before sizing anything: gold helps against a falling dollar, and would add to the losses if the risk were a rising dollar. Picking a hedge that loses at the same time as the thing it protects is the most common hedging error.

How producers and businesses hedge gold

The original users of gold hedging are businesses with real gold exposure:

• A miner worries that gold will fall before its metal is sold, so it sells futures or forward contracts for part of its future output. If gold falls, the futures gain offsets the lower sale price; if gold rises, the futures lose and the miner gives up some upside. The point is certainty, not profit.
• A jeweller or refiner worries that gold will rise before it buys metal for orders already priced, so it buys futures to lock in the cost.
• An importer in a country like India faces two risks at once, the gold price and the rupee, and may hedge both.

Worked example (hypothetical): a refiner has agreed to deliver 100 ounces of finished gold in three months at a price based on today’s 4,000. It will buy the metal later. To lock in the cost, it buys one GC future at 4,030 (spot plus carry). In three months gold is at 4,200. The refiner pays $20,000 more for its metal than planned, and the future, now near 4,200, has gained about $17,000. The hedge offset most of the move; the $3,000 difference is roughly the carry it paid. Our gold futures lesson explains the basis behind that gap.

A trader’s position is not like this. A trader has no metal to protect. That is why, for traders, “hedging” usually means controlling the risk of the trade itself.

What about silver hedging?

Silver producers and users hedge the same way, with silver futures or forwards. For a gold trader, silver is not a hedge: it moves in the same direction as gold on most days and by more, so adding silver to a gold long increases risk. Silver only reduces gold risk when it is held in the opposite direction, which turns the position into a ratio trade.

How traders use the hedge evidence

Even if you never hold gold as a hedge, the evidence above is useful for trading gold:

• Dollar days: a big dollar move is the most reliable short-term headwind or tailwind for XAUUSD. Check the dollar before fading a gold move.
• Crash days: don’t assume gold will rise on a stock crash. The first move is often down, with stocks, as margin calls hit. Plan for both.
• Inflation releases: high inflation that raises rate expectations is often bad for gold on the day. Our CPI-day lessons cover the reaction.
• Haven flows over weeks: after the first panic, gold has often recovered and rallied. Swing traders watch for the turn rather than buying the first day.

Base rates in one place

• Dollar: gold moved opposite to the dollar on about two days in three over two years; correlation -0.38.
• Inflation: 9.0% a year vs 3.9% CPI from 1970 to 2025; beat inflation in 7 of 14 high-inflation years.
• Stocks: up on 5 of 10 days the S&P fell 2% or more; correlation 0.33.
• April 2025: gold -5.6% in the crash week, then +14.6% from the low.
• Portfolio: 10% gold cut the two-year S&P drawdown from 18.9% to 17.0%.

Pro nuance

• Hedges are regime-dependent. Gold’s link to stocks and the dollar changes over time. The site’s correlation tool shows 90-day correlations drifting by large amounts within a year. A hedge sized on last year’s correlation can be wrong this year.
• Real yields are the hidden variable. Most of gold’s failures as an inflation or crash hedge came when real yields rose. Watching them explains more than watching inflation.
• Currency matters. A hedge that works for a US-dollar investor may not work for one in rupees, yen or euros. Measure gold in the currency you spend.
• Cost of carry. Holding gold through futures or CFDs costs financing; holding physical gold costs storage and wide dealer spreads. A hedge you can’t afford to hold for years isn’t a long-term hedge.
• Silver is not a gold hedge. Silver usually moves with gold, by more. It increases metal risk rather than reducing it, unless you hold it short against gold on purpose.

Common mistakes

• Expecting gold to rise on the first day of a crash. It fell in the first week of April 2025, in 2008 and in March 2020.
• Buying gold “for inflation” without watching rates. In 2021 and 2022 inflation was high and gold fell.
• Opening an opposite position instead of using a stop. It locks the loss and doubles the costs.
• Using a hedge you don’t understand. Options, ratio trades and correlated positions each have their own failure points.
• Judging a hedge on one episode. Every claim in this lesson has counter-examples; look at the base rates.
• Calling a speculative position a hedge. If you are trading gold, the risk is the trade itself; hedge it with size and stops.

Quick checklist

1. What exactly am I hedging: the dollar, inflation, stocks, a currency, or a single trade?
2. Over what time frame, and does gold’s record fit that time frame?
3. What happens to my hedge in the first days of a crash, when gold often falls too?
4. What do real yields and the central bank imply for gold right now?
5. What does the hedge cost to hold (financing, storage, premiums)?
6. For a trade: is my stop and size the main protection, rather than an opposite position?
7. If I use a correlated position, what is my plan if the correlation breaks?

Sources

• US Bureau of Labor Statistics: Consumer Price Index (CPI-U; the site’s series is FRED CPIAUCSL)
• Federal Reserve Bank of St. Louis (FRED): CPIAUCSL
• World Bank: Commodity Markets (Pink Sheet) (monthly gold prices before 2001 in the site’s history file)
• Federal Reserve: monetary policy
• Our own data: the site’s gold history file (yearly gold 1968-2026, inflation-adjusted series in September 2026 dollars), COMEX gold futures daily closes via Yahoo Finance (switch-adjusted for changes between dates), S&P 500 and ICE dollar index daily closes via Yahoo Finance (4 November 2024 to 3 November 2026), and the site’s 250-day correlations to 4 November 2026. The 2008, 2020 and 2022 figures are from our crash and inflation studies linked above.

Educational content only, not financial advice. Trading gold with leverage carries a high risk of loss.

Frequently asked questions