What If I Invested $1,000 in Gold Ten Years Ago?

Somewhere in a filing cabinet in America there is a receipt from the autumn of 2016. A thousand dollars, one line item, a few gold coins in a padded envelope. The buyer probably told almost nobody, put the coins somewhere safe, and got on with a decade of life.

Here is what that receipt is worth now. Gold finished September 2016 at roughly $1,319 an ounce, so $1,000 bought about 0.758 of an ounce. At today’s price of roughly $4,390, that same slice of metal is worth about $3,330. Call it 3.3 times the money, a gain of about 233%, or roughly 12.8% a year. The coins never sent a statement, and nobody had to be paid to manage them.

Now the part most articles quietly skip. Over that identical window, $1,000 in a plain S&P 500 index fund with dividends reinvested would be worth closer to $4,100, something like 15% a year. Stocks won the decade, and not by a hair. If your only question is which line on the chart went up more, the honest answer is equities, and anyone who tells you otherwise is selling something.

So why does the gold buyer seem so unbothered? Because the number on the chart was never the whole transaction. One of these two owns a claim on a system. The other owns a metal disc in a safe that needs nobody’s promise to remain what it is. And there is a third person we have not mentioned: the one who left the $1,000 in the bank.

Ten Years of Spot Prices, Dividends, and Quietly Vanishing Purchasing Power

Two terms deserve defining. The spot price is the wholesale price for one troy ounce of metal for immediate delivery, the number quoted on financial sites. The premium is what you pay above spot to hold a coin or bar, covering minting, distribution, and the dealer’s margin. Nobody buys physical metal at spot, and any calculator pretending otherwise is running a simulation, not a purchase.

The decade itself was strange. For years gold went nowhere while stocks compounded quietly and the Federal Reserve kept interest rates on the floor. Then came a wave of central bank buying and a repricing that carried the metal above $4,000 an ounce for the first time in October 2025. Gold closed February 2026 near $5,279, fell to about $4,012 by the end of June, and sits today around $4,390, still roughly 17% below that high.

That volatility is real, and pretending it away does the reader no favors. But volatility is not the same thing as permanent loss, and here the third person in our story matters. The saver who left $1,000 in a “risk-free” account still has roughly $1,000, plus whatever crumbs the bank passed along. What they no longer have is the ability to buy what $1,000 bought in 2016. Ask anyone who has filled a grocery cart lately. The dollar’s decline is not a forecast; it is a receipt.

Four Factors That Decide What a Thousand Dollars in Bullion Actually Returned

The premium you paid is a real cost, and it is smaller than you fear

Say you paid a 5% premium in 2016. Your $1,000 bought about 0.722 of an ounce instead of 0.758, worth roughly $3,170 today rather than $3,330. The premium cost about $160 on a $2,300 gain, trimming the annualized return from roughly 12.8% to about 12.2%.

That is the whole story on premiums. A one-time entry cost divided across every year you hold, which is why the long-term buyer barely feels it. Where premiums genuinely punish you is in churning, which is not what physical bullion is for.

Dividends do work that gold cannot do

The S&P 500 advantage over this decade was not magic. It was earnings and dividends, cash thrown off by real companies selling real products, then reinvested. Gold has no cash flow, pays no dividend, and issues no earnings report. Anyone expecting metal to behave like a productive business has misunderstood the purchase.

Counterparty risk is the line item nobody quotes

Counterparty risk is the chance the other party to your arrangement fails to deliver. Every share, bond, ETF, and bank deposit carries some. A stock certificate is a claim on a company, held through a broker, cleared through a system, denominated in a currency managed by people who have never volunteered to manage it carefully. 

A gold coin in your possession is a claim on nobody. It cannot be diluted by a share issuance, frozen by a platform outage, or voted away by a committee. That safety is not a return you can chart, which is exactly why it never appears in the ten-year comparison.

Storage is a real cost, and a modest one

Physical metal has to live somewhere: a home safe, a bank box, or a professional depository. Depository storage typically runs a fraction of a percent per year, comparable to what many investors pay in fund fees without blinking. The honest comparison is not “storage versus free.” It is storage against the fees, spreads, and platform risk attached to every paper alternative.

A Simple Framework for Reading Any Backward-Looking Gold Number

If you are running this math to decide what to do next, a few rules keep you honest.

Change the decade and the answer flips. Gold has beaten the S&P 500 over the last five years, roughly 20% a year against 13 to 14%. Over 25 years it is not close: gold traded near $270 an ounce in the fall of 2000, about sixteen times your money since, roughly 11% a year against roughly 8% for stocks with dividends reinvested. The two take turns leading for a decade at a time, so picking a start date decides the winner in advance.

If you want growth, own equities. If you want an asset that cannot default, own metal. Those are different jobs. Portfolio diversification means holding things that fail at different times, not stacking things that depend on the same monetary plumbing.

Buy in a form you can identify and sell easily. Widely recognized gold coins and standard bars from known refiners trade with tight spreads. Obscure products do not.

And if you cannot leave the money alone for five years, do not buy physical gold at all. The entry premium and the metal’s habit of moving sharply in either direction will chew up a short holding period.

Common Concerns From People Running This Math

“Stocks won, so why bother?” Because the decade you are measuring is not the decade you will live through. The Federal Reserve now carries a balance sheet and a debt load no prior generation of policymakers had to manage, and central banks abroad have been accumulating gold rather than lending more freely to Washington. 

Stocks are a bet on corporate earnings inside a stable monetary order. Gold is what you own in case that order proves less permanent than the politicians, the central bankers, and the paper pushers insist. Owning both is the point.

“Doesn’t the premium ruin the return?” No, and the arithmetic above shows why. A 5% entry premium cost that 2016 buyer about half a percentage point a year. Set that against a decade of fund expenses and taxable distributions on the paper side and the picture is less lopsided than it looks.

“Am I too late?” Gold sits roughly 17% below its February high, an odd position for a supposed bubble. Nobody knows the next move, and anyone who claims to should be ignored on principle. What can be said is that buyers of physical metal are not trying to catch a top. They are trying to hold purchasing power across decades, and they count results in ounces, not in Federal Reserve notes.

Conclusion

A thousand dollars in gold ten years ago became roughly $3,330. The same thousand in an index fund became roughly $4,100. Both beat the thousand left in the bank so thoroughly the comparison is almost unkind. The stock investor made more money over that particular decade. 

The gold buyer bought something else entirely: an asset with no counterparty, no dividend, and no expiration, bought at a real price with a real premium and stored at a real cost. Ten years on, the metal is still exactly what it was, which was always the entire proposition.