Is there a safe rule-of-thumb percentage for gold?

You'll see specific percentages quoted everywhere, often by firms that sell gold. Treat them as someone else's starting point, not a fit for your situation. The right figure depends on your time horizon, what you already own, why you're buying, and how much volatility you can hold without selling. Use the framework above to reason to your own number rather than adopting a stranger's.

Is owning some gold better than owning none?

For many diversified investors, a modest position adds something equities and bonds can't — a near-zero correlation to stocks and a hedge with no counterparty (World Gold Council). Whether you should hold any, and how much, still depends on your circumstances. This is education, not a recommendation to buy.

Why do central banks hold so much gold?

Because it's an asset that isn't anyone else's liability and doesn't depend on another country's currency or solvency. Central banks collectively hold about 17% of all gold ever mined and have been net buyers in recent years (World Gold Council). Their motives — reserve diversification and monetary independence — overlap with an individual's, but their scale and goals are not yours.

Does gold protect against inflation?

Historically gold has held purchasing power across major inflationary regimes, including the 1970s, which is why it's discussed as an inflation hedge. "Historically" is the operative word — it's not guaranteed in any given year, and gold can fall while inflation runs. Treat it as one tool among several, not a promise.

What's the downside of owning gold?

It pays no income, it can underperform for a decade or more (as it did after the 1980 peak), and it's volatile month to month. Those are real costs, and they're the main reason to size a position deliberately rather than overload on it.

Physical gold or a gold ETF — does the form change how much I should own?

The form changes cost, liquidity, control, and tax treatment, which can influence how much you're comfortable holding. A deeper comparison is in physical vs paper gold. The decision on how much still flows from your own variables, not from the wrapper you choose.

How much gold should I own? A framework

By Michael, founder of Capital Fortress. PhD in Finance. 25+ years trading institutional commodities on the CBOT and CME floors. Four businesses built to seven- and eight-figure revenues. Active trader across forex, equities, gold, and commodities. Michael writes for investors who want a framework, not headlines, and a plan, not opinions. More about the author →

The honest one-sentence answer: there is no universal right number, and anyone who hands you a single percentage without knowing your situation is selling something. What I can do is give you the part nobody can outsource to a chart — understanding what gold actually does in a portfolio, what it doesn't do, and the variables that move the number up or down for you. After that, the percentage is yours to set.

I've watched this question get answered badly for 25 years. I traded commodities through the CBOT/CME floor era, sat through the 2008 credit freeze, the 2020 crash, and the 2022 inflation shock. The dealers who want to sell you coins will tell you a fixed percentage. The advisor who has never held a hard asset might tell you to skip it entirely. Both are guessing about you. Let's replace the guess with a framework.

What gold actually is, before you decide how much to hold

Gold is one of the only assets you can own that is nobody else's liability. A stock is a claim on a company. A bond is a promise from a borrower. A bank deposit is an IOU from the bank. Each of those depends on a counterparty staying solvent and honest. Physical gold doesn't. That single property — no counterparty — is the reason it has been treated as money for thousands of years and why it still sits on central-bank balance sheets today.

That's also why the monetary history matters. For most of the 20th century the U.S. dollar was tied to gold at a fixed $35 an ounce. That ended on August 15, 1971, when President Nixon suspended dollar-to-gold convertibility and closed the gold window (Federal Reserve History). Once gold was free to trade, it did something instructive during the inflationary 1970s: it ran from $35 to roughly $850 an ounce by January 1980 (World Gold Council price data). Gold behaved like an escape valve when confidence in paper money cracked.

What gold does well

Here is the case for owning some, grounded in evidence rather than enthusiasm.

It diversifies in a way few assets can. Over the long run gold has shown a low, near-zero correlation with equities, and it tends to decouple from stocks precisely when stocks are selling off (World Gold Council — Relevance of Gold as a Strategic Asset). A portfolio holds up better when its pieces don't all fall together. Gold is one of the few that has historically zigged when equities zagged. If you want the wider picture of which assets behave this way, I covered it in safe-haven assets.

It has held purchasing power across regime changes. Through the end of Bretton Woods, the 1970s inflation, and modern crises, gold has been a store of value when currencies were debased. That's the same reason it sits at the center of any serious conversation about an inflation hedge.

It is liquid and universally recognized. You can sell gold anywhere on earth. That portability is part of why it functions as financial insurance.

What gold does NOT do — read this before you size it

A framework that only lists the upside is a sales pitch. Gold has real costs, and pretending otherwise is how people end up over-allocated and disappointed.

It pays you nothing. No dividend, no interest, no coupon. A share of a profitable business or a bond can compound; a bar of gold just sits there. Its entire return depends on the price someone else will pay later.
It can underperform for a very long time. After that 1980 peak, gold spent roughly two decades going sideways-to-down in real terms before its next major run. Anyone who bought the top in 1980 waited a generation to get whole. The 1970s climb wasn't smooth either — gold ran up, then gave back a large share of the gain mid-decade before resuming, so even its best decade punished anyone who needed the money at the wrong moment.
It is volatile. As I write this on June 9, 2026, gold trades around $4,300–$4,360 an ounce, up roughly 30% over the past year but down close to 9% in the prior month alone (Fortune, Trading Economics). That is normal for this asset. If a 9% monthly swing would make you sell in a panic, that tells you something about your own number.

Gold is insurance, not a growth engine. You don't judge insurance by how much it returns in a calm year. You hold it for the years that aren't calm.

What some institutions and central banks do — as context, not instruction

People reach for "what do the pros hold?" as a shortcut to a number. It's useful context, so long as you remember it is context, not a prescription for your household.

Central banks have been steady, large buyers. They collectively hold around 17% of all the gold ever mined, with official reserves topping roughly 36,520 tonnes at the end of November 2025 (World Gold Council — Gold Reserves by Country). They bought about 863 tonnes in 2025 — below the 1,000-plus tonnes of each of the prior three years, but still well above the 2010–2021 average of roughly 473 tonnes a year (World Gold Council — Gold Demand Trends 2025). In the World Gold Council's 2025 survey, 95% of responding central banks expected global official gold reserves to keep rising over the following year (WGC Central Bank Gold Reserves Survey 2025).

Worth knowing: when you see currency-reserve breakdowns, gold isn't in them. The IMF's COFER data tracks the currency composition of foreign-exchange reserves; gold is held and reported separately as an official reserve asset (IMF COFER). So a central bank holding "X% gold" and an investor's portfolio percentage aren't measuring the same thing.

As for portfolio research, the World Gold Council's own work has found that modest gold positions historically improved risk-adjusted returns across different portfolios and regions (WGC strategic-asset research). I'm pointing to the direction of that research — not putting a number on your portfolio. A nation managing reserves and a 38-year-old saving for retirement have almost nothing in common. Borrow the logic, not the number.

The variables that actually set your number

This is the work no article can do for you, because the answer lives in your circumstances. Run through these honestly and the right range for you starts to come into focus.

Time horizon

A long horizon lets you ride out gold's flat decades and lean harder on growth assets that compound. A shorter horizon, or money you'll need soon, changes how much volatility you can stomach.

What you already own

Gold's value is mostly about what it isn't. If your wealth is concentrated in U.S. equities, a small gold position adds genuine diversification. If you already hold real estate, foreign assets, and other uncorrelated pieces, you may need less. Look at the whole portfolio, not gold in isolation — the same principle behind building a recession-proof portfolio.

Why you want it

Are you buying insurance against a currency or systemic shock, or chasing a price you saw on the news? Insurance-sized thinking and speculation-sized thinking produce very different numbers, and only one of them survives a bad month.

Your tolerance for doing nothing

Gold can sit dead for years. If holding an asset that "does nothing" while stocks rip higher would drive you to sell at the wrong time, your real capacity for gold is lower than a spreadsheet suggests.

How you'd hold it

Physical metal, an ETF, and a tax-advantaged structure behave very differently on cost, liquidity, and control. That choice interacts with how much you hold — I walk through the trade-offs in physical vs paper gold and the broader mechanics in gold investment. And if you're weighing the white metal alongside it, gold vs silver covers why the two don't play the same role.

So, how much should you own?

Put the variables together and you'll land on a range that fits your horizon, your existing holdings, your reason for buying, and your temperament — not a number I handed you. Some people, after working through this, conclude a small insurance-sized sliver is right. Others with different circumstances land higher or lower. Both can be correct, because "correct" is defined by your situation, not by a headline percentage.

What I'd steer you away from is the two lazy answers: zero (you give up real diversification and a hedge with a thousand-year track record) and "whatever the dealer said" (a number set by someone whose income depends on you buying more). The grown-up answer is a deliberate position you can hold through a bad month without flinching.

This is exactly the kind of decision the broader Capital Fortress SAFE framework goes deeper on — applying the role of hard assets under different economic conditions rather than freezing on one fixed figure. The thinking here is the foundation; SAFE builds the application on top of it.

Gold over the long run: large gains, long flat stretches

Source: World Gold Council price data. Gold left the $35 official price when the gold window closed in 1971, peaked near $850 in January 1980, then went roughly two decades sideways into the early 2000s before the run to about $4,300 by 2026.

If you want to see how a gold position sits inside your actual asset mix rather than in the abstract, Command Center lets you map your holdings and watch how diversification and concentration shift as you change them — useful for pressure-testing your own number before you commit a dollar.

How Much Gold Should I Own? A Framework, Not a Number