Is a Gold IRA a Good Investment? Pros, Cons and 50 Years of Data (2026)

That answer will not satisfy anyone who has seen an advertisement promising that gold "protects your retirement" - or anyone who has been told gold is a "pet rock" that only fools buy. Both views leave things out. This guide lays out the evidence on both sides, in plain English, with the numbers shown so you can check them yourself.

Here is what you will find below:

• The data. Year-by-year gold prices from 2015 to 2025, the record run to about $5,595 an ounce in January 2026, the 29% drop that followed, and how gold compares with stocks over three different long periods.
• The worked math. What $10,000 would have grown to in gold and in stocks, how long gold took to recover from its worst crashes, and exactly how dealer markups and yearly fees change your break-even point.
• Both cases. A fair list of the real advantages, and an equally detailed list of the real drawbacks - the ones sales calls tend to skip.
• The alternatives. How a gold IRA compares with a gold ETF, coins in your own safe and gold-mining stocks.
• A self-check. Profiles of who a gold IRA can make sense for, a checklist of questions to ask yourself, and answers to the six questions we hear most.

If you are still unsure what a gold IRA actually is, start with our short explainer What Is a Gold IRA? or the full Gold IRA guide, then come back. This article assumes you know the basics - that a gold IRA is a self-directed IRA holding IRS-eligible bullion through a custodian and an approved depository - and focuses on one question: is it worth it?

How we wrote this

• Every figure is dated. Prices and fees are as of October 19, 2026 unless stated otherwise. Gold moves every minute.
• We show our arithmetic. Where we calculate something - growth of $10,000, a break-even, a percentage drop - we show the inputs so you can redo it. Calculated figures are approximate and before taxes unless we say so.
• Past performance is not a forecast. Nothing in the history below tells you what gold will do next year or next decade. We use history to show the range of outcomes, not to predict one.
• This is education, not advice. Whether gold belongs in your retirement plan depends on your income, other savings, health, taxes and goals. A fee-only financial adviser - one paid by you, not by commissions on what you buy - is the right person to make that call with you.

"Is gold a good investment?" is really six questions in one. An asset can be excellent at one job and poor at another. A 30-year-old building wealth and a 70-year-old living off savings need very different things, so before looking at gold specifically, it helps to be clear about what you are asking it to do.

Put like that, the honest summary is: gold is not built to be a growth engine or an income source. Its strongest case is as a diversifier - something that may hold up, or even rise, when other parts of your savings are falling. Whether that job is worth the price you pay in a gold IRA is the question the rest of this guide tries to answer.

Why retirement changes the scoring

The same asset can be a good fit at 45 and a poor one at 70. Three things shift as you approach and enter retirement:

1. Time horizon shrinks. At 45 you might wait out a 10-year slump. At 70, a 10-year slump is a large share of the time you will be drawing on the money.
2. Withdrawals begin. Once you are taking money out, a falling asset forces you to sell more units to raise the same cash. Traditional IRA owners must also take required minimum distributions (RMDs) from age 73 (75 for people born in 1960 or later), whether or not prices are good. Our gold IRA taxes and RMDs guide explains how that works with metal.
3. Income matters more. Many retirees want their savings to produce cash flow. Gold cannot do that; you have to sell some.

None of this rules gold out. It means that, for a retirement saver, the bar is higher: gold has to earn its place mainly through diversification, and the cost of holding it has to be low enough not to eat the benefit.

A simple way to frame your own question

Instead of "Is gold good?", try asking: "If I moved X dollars of my retirement savings into physical gold in an IRA, paying these specific costs, would my overall plan be more likely or less likely to meet my needs?" That framing forces you to name the amount, the costs and the purpose. We come back to it in the checklist near the end.

A share of stock is a claim on a company's future profits, and a bond is a promise of interest. Gold is neither. It is a metal whose price is set by what buyers around the world will pay for it at any moment. To judge whether it suits you, it helps to understand who those buyers are and what moves them. Economists argue about the exact weights, but most agree on these main drivers.

1. Real interest rates

A "real" interest rate is the interest you earn on safe savings - such as government bonds - minus inflation. Because gold pays no interest, holding it has an opportunity cost: the income you give up by not owning a bond or a savings account instead.

• When real rates are high (say, bonds paying well above inflation), that opportunity cost is large, and gold has tended to struggle.
• When real rates are low or negative (bonds paying less than inflation), giving up interest costs little, and gold has tended to do better.

This is a tendency, not a law. The relationship has weakened at times, and it does not tell you what rates will do next.

2. The US dollar

Gold is priced in dollars worldwide. When the dollar weakens against other currencies, gold becomes cheaper for buyers using euros, yen or rupees, which can lift demand and the dollar price. When the dollar strengthens, the reverse can happen. Again: a tendency, frequently overridden by other forces.

3. Central-bank buying

Central banks - the institutions that manage a country's currency and reserves - hold gold as part of their reserves. In recent years they have been buying at historically high levels, which has been one of the most discussed forces behind the rally. Reporting on the market describes record central-bank buying, and total gold demand reached a record 1,231 tonnes in the first quarter of 2026. The World Gold Council's Gold Demand Trends report for full-year 2025 is the best place to see the breakdown by buyer type; it also records a record annual average price of $3,431 for 2025, with 53 record LBMA PM prices set during the year.

The important point for a retirement saver: central-bank demand can change. Their buying decisions depend on politics, reserve policy and prices, and a slowdown in official buying would remove one of the supports under recent prices.

4. Crisis and fear demand

When investors are worried - about wars, banking problems, government debt, elections or sharp stock-market falls - some move money into gold. This "fear trade" can push prices up quickly. It can also reverse quickly when the fear fades, which is one reason gold's price can swing so sharply.

5. Investment flows

Investors buy gold through exchange-traded funds (ETFs), bars and coins, and futures contracts. Big flows into or out of gold ETFs can move the price for weeks or months. A wave of buying by people who saw prices rising - "momentum" buying - can carry prices above what other factors justify, and a wave of selling can do the opposite.

6. Jewelry and industrial demand

A large share of the gold mined each year ends up as jewelry, especially in countries with strong cultural traditions of gold-giving. Some is used in electronics, dentistry and other industries. Jewelry buyers tend to be price-sensitive: when gold gets very expensive, they buy less, which can soften demand at high prices.

7. Mine supply

New mine output adds to supply only slowly; new mines take many years to develop. Recycled gold (old jewelry and scrap) responds faster, rising when prices are high.

What this means for you

Notice that none of these forces is a steady engine of growth like company earnings. Gold's price is the net result of many buyers with very different motives, and it can move far and fast in either direction. That is the core reason gold is volatile - and the reason no one, including the people selling it, can tell you where it will be in five years.

You can see the official benchmark price history from the LBMA (London Bullion Market Association), which publishes the twice-daily LBMA Gold Price used by much of the industry.

Long-run averages hide how bumpy the ride was. Here is gold year by year over the last decade-plus, using year-end prices per troy ounce. We show two "change" columns: the yearly change as reported in our source data, and the change we calculate from the year-end prices in the table. They do not always match, because published yearly changes are often based on slightly different closing prices or benchmark fixes. Neither is "wrong" - but you should know that even simple price history varies by source.

Sources: year-end prices via JM Bullion; reported yearly changes from published annual returns. Calculated column is our arithmetic, for example 2016: $1,152 / $1,061 - 1 = +8.6%. For 2024 and 2025 we used the midpoint of the reported ranges.

What the table shows

• Four down or flat years out of eleven. 2015, 2018, 2021 and 2022 were losing or roughly flat years. Someone who bought at the end of 2020 at about $1,899 was still slightly behind two years later.
• Most of the gain came in a few bursts. 2019-2020 and 2024-2025 did most of the work. In the years between, gold went sideways.
• The decade was unusually strong. From about $1,061 at the end of 2015 to about $4,316 at the end of 2025, gold roughly quadrupled. The math: $4,316 / $1,061 = about 4.07 times, which works out to roughly 15% a year compounded over 10 years (4.07 to the power of 1/10 = about 1.151). That is far above gold's long-run average of about 8.8% - because 2015 was near the bottom of a four-year slide and 2025 was a record year. Pick different end points and you get a very different number.

Annual average prices tell a calmer story

Year-end prices capture one day. Annual averages smooth out the noise:

Notice the three years from 2020 to 2022: the average price barely moved ($1,770 to $1,800), even though those years included a pandemic and a surge in inflation - exactly the kind of events gold is supposed to respond to. We return to that point in the inflation section.

The 2025 average of $3,431 was about 44% above 2024's $2,386 ($3,431 / $2,386 - 1 = 0.438). That is a big jump in a single year for an asset many people buy for "stability."

The last three years are the backdrop to almost every gold IRA advertisement you see today, so they deserve a close look.

The rise

• End of 2023: about $2,063 an ounce.
• End of 2024: about $2,610-$2,624, up roughly 27%.
• End of 2025: about $4,310-$4,323, up roughly 65% in a single year.
• January 29, 2026: a record of about $5,595 an ounce intraday (Reuters). The LBMA morning fix that day was $5,501.70 and the afternoon fix $5,405.00.

From the end of 2023 to the January 2026 peak, gold rose about 171% ($5,595 / $2,063 - 1 = 1.71). Even from the end of 2025, the extra climb to the peak was about 30% in four weeks ($5,595 / $4,316 - 1 = 0.296).

The fall

• January 30, 2026: spot fell 3.9% to about $5,183 in one day.
• Late June 2026: the LBMA fix hit a low of about $3,994.50 - roughly 29% below the peak ($3,994.50 / $5,595 - 1 = -0.286). The 52-week low in spot trading was around $3,900.
• October 19, 2026: about $4,170 (Kitco).

Our arithmetic on spot prices, for example $4,170 / $5,595 - 1 = -25.5%. These figures ignore dealer markups and fees, which would make every row worse for a gold IRA buyer - by several percentage points or more. See how fees change the math.

Why this matters

Two lessons stand out. First, timing dominates short-term results: the same metal produced a doubling for one buyer and a 25% loss for another, depending only on the purchase date. Second, a 29% fall in five months is a large, fast drop - bigger than many stock-market corrections. Anyone who bought near the top in a gold IRA expecting "stability" got the opposite.

That does not mean gold was a bad buy in 2026 or that it will not recover; it may, and it may not. It means the honest description of gold is "volatile asset that sometimes diversifies" - not "safe store of value."

What forecasters said - and why to be cautious

Big banks publish gold forecasts. Earlier 2026 figures included an average of about $4,325 (ING), $4,900 (Goldman Sachs) and $5,000 (Bank of America); by mid-to-late 2026 several banks had cut theirs - for example Goldman Sachs to about $4,650 for end-2026 and Bank of America to a 2026 average of about $4,360 (see our gold price history for sources). Note how widely they differ, and remember that forecasts made before January 2026 did not predict either the spike to $5,595 or the fall to under $4,000. Treat any price forecast - from a bank or a gold dealer - as an opinion, not information.

Gold's price has floated freely since the early 1970s, when the US ended the link between the dollar and gold. That gives us roughly half a century of data. Here is how gold compares with the S&P 500 (with dividends reinvested) over three commonly cited periods:

"Nominal" means the plain dollar return. "Real" means the return after subtracting inflation - what your money could actually buy. All of these are before taxes, fees and dealer costs, and all are for the gold price, not for a gold IRA, which would trail the price because of costs.

Why the start date changes the answer

Look at the last row. Over 2000-2025, gold beat stocks. Over 1971-2025 and 1995-2025, stocks beat gold. How can both be true?

Because the starting point matters enormously when an asset moves in long cycles:

• January 2000 was the peak of the dot-com stock bubble. Stocks then had two major crashes in the next nine years. Meanwhile, gold was near the bottom of a 20-year slump (it hit about $252 in August 1999). Starting there flatters gold and penalizes stocks.
• 1995 started with stocks reasonably priced and gold already in its long slide. That flatters stocks.
• 1971 captures gold's huge 1970s run and its 1980-1999 collapse, plus every stock boom and bust since - which is why it is often treated as the fairest long view.

The practical lesson: whenever someone quotes you a gold return, ask "from when to when?" A salesperson can make gold look brilliant or a stock fan can make it look terrible simply by choosing the dates. The longest period in our data - since 1971 - shows stocks ahead by about 2.4 percentage points a year, and that gap compounds into a very large difference over decades, as the next section shows.

Why stocks have usually won over long periods

The basic reason is simple. A stock index represents a share in thousands of businesses that sell goods and services, earn profits, reinvest them and pay dividends. As the economy grows, those profits tend to grow. Gold produces no profit and pays no dividend; its return comes only from price changes. Over very long periods, an asset that generates cash has had a structural advantage over one that does not.

That does not mean stocks always win - 2000-2025 shows otherwise - or that gold is useless. It means gold's role, if it has one in your plan, is usually as a complement to stocks and bonds rather than a replacement. For a broader look at the different ways to own gold for retirement, see how to invest in gold for retirement.

Percentages per year are hard to feel. Here is what those reported annual rates would turn $10,000 into, using compound growth. The formula is: ending value = $10,000 x (1 + annual rate) to the power of the number of years.

Since 1971 (about 54 years, nominal)

A 2.4-point annual gap became a roughly 3-to-1 difference in ending wealth. We used 54 years (1971 to 2025); counting 55 years would raise both numbers by about 9-11% but leave the comparison unchanged. These are nominal dollars - inflation over half a century means a dollar in 2025 buys far less than a dollar in 1971.

1995-2025 (30 years)

The "real" rows are the most meaningful for a retirement saver, because they measure what the money can actually buy. Gold multiplied purchasing power by almost 5 times in 30 years - a solid result - while stocks multiplied it by about 11 times.

January 2000 - August 2025 (about 25.6 years, nominal)

Same two assets, opposite result: gold came out ahead by about $56,000 on a $10,000 start. Someone who retired in 2000 with a meaningful slice in gold would have been glad of it.

Important caveats on all three tables

• These are approximate. They apply a single average rate to every year. Real returns arrive unevenly, and the published annual rates are themselves rounded.
• No costs are included. A gold IRA buyer pays a dealer markup on the way in, a spread on the way out and custodian and storage fees every year. A stock index fund investor pays a small expense ratio. Costs would reduce both columns, but gold's more.
• No taxes are included. Inside an IRA, growth is tax-deferred (traditional) or tax-free (Roth) either way, so the comparison holds within an IRA. Outside an IRA the tax treatment differs - see our taxes and RMDs guide.
• Past results do not predict future ones. Nothing here tells you which asset will do better from 2026 onward.

A "drawdown" is the fall from a peak to the lowest point before a new peak. For a retirement saver, drawdowns matter at least as much as average returns, because a deep drop at the wrong moment can permanently damage a plan. Gold has had three big ones in modern history:

The math for the last column: to recover from a fall of X%, you need a gain of 1 / (1 - X) - 1. For a 70% fall: 1 / 0.30 - 1 = 2.33, or +233%. For 45%: 1 / 0.55 - 1 = +82%. For 29%: 1 / 0.71 - 1 = about +41% (using the exact prices, $5,595 / $3,994.50 - 1 = +40%). Losses hurt more than gains help - a basic fact of investing that applies to every asset.

How long recovery took

• 1980 peak: gold was still about 70% below its January 1980 high almost 20 years later, in August 1999. By most accounts it did not regain $850 in nominal terms until around 2008 - close to three decades after the peak. (Our data confirms it was above $850 by the end of 2015, when it closed at $1,061.) After inflation, the recovery took far longer still, because the cost of living kept rising throughout those years.
• 2011 peak: on our year-end figures, gold closed 2019 at $1,517 and 2020 at $1,899 - still below the 2011 high of $1,921. The first year-end close above it was 2023 ($2,063), about 12 years after the peak. Prices briefly topped the old high during 2020, but a buyer at the 2011 peak would have spent most of a decade under water.
• 2026 peak: as of October 19, 2026, gold is about $4,170, roughly 25% below the January record. Nobody knows when, or whether, it will regain $5,595.

What a drawdown means for someone aged 65

Imagine you are 65, retired, and you need to raise $10,000 from your gold IRA for living costs or a required minimum distribution. Compare selling at the 2011 peak with selling near the 2015 low:

At the low, you would have to sell about 83% more metal to raise the same cash - and those ounces are gone for good, so they cannot benefit from any later recovery. (Real sales would also be slightly below spot after the dealer's buy-back spread.) This is called sequence-of-returns risk: a loss early in retirement, while you are withdrawing, does more lasting damage than the same loss later or before you retire.

Three practical points follow:

1. Do not hold money you will need in the next few years in gold. Many planners suggest keeping several years of withdrawals in cash or short-term bonds so you are never forced to sell a volatile asset at a bad time.
2. A 12-year or 28-year recovery is longer than many retirements' "spending window." At 65, the 1980 experience would have lasted until you were in your 90s.
3. RMDs do not wait for a recovery. If most of a traditional IRA is in gold, you must still take your required distribution each year - either by selling metal or by taking the coins or bars "in kind" and paying tax on their value. Our gold IRA rules guide covers the mechanics.

Stocks have drawdowns too, of course, sometimes deeper and faster than gold's. The point is not that gold is uniquely risky, but that it is not the low-risk asset its marketing often implies.

"Gold protects you from inflation" is probably the most common reason people give for buying it. The evidence is genuinely mixed, and the honest answer depends on the time frame.

The case that it is - over very long periods

Over 1995-2025, gold returned about 8.1% a year nominal and about 5.4% a year after inflation (Kiplinger). Those two figures let us back out the average inflation rate the source used:

1.081 / 1.054 - 1 = about 2.6% a year

So over those 30 years, gold beat inflation by a wide margin - its purchasing power grew by roughly 5.4% a year. That is consistent with the popular idea that gold holds its value over the long run. The same is true of the 1971-2025 period as a whole: gold's ~8.8% nominal return was well above the general rate of price increases over that time. In the 1970s, when inflation was high, gold rose from the old official price of $35 an ounce to $850 by January 1980.

The case that it is not - over the periods that matter to a retiree

The problem is that "the long run" for gold can be longer than your retirement. Consider:

• 1980-1999. Consumer prices rose every year during the 1980s and 1990s, yet gold fell about 70% in nominal terms, from $850 to $252. Someone who bought gold in 1980 to protect against inflation lost most of their purchasing power and waited decades to get it back. This is the clearest counterexample to the "inflation hedge" claim.
• 2021-2022. The US experienced its sharpest burst of inflation in about four decades. Gold's reported yearly change was -5.8% in 2021 and about +0.4% in 2022. Its annual average barely moved: $1,799 in 2021 and $1,800 in 2022. In the very period when an inflation hedge was most needed, gold lagged rising prices.
• 2024-2025. Gold then soared (+27% and +65%) while inflation was cooling from its peak. Gold's biggest gains came after the inflation scare, not during it.

Why the link is loose

Recall the drivers from earlier. Inflation is only one of them, and it often works through real interest rates. In 2022, for example, central banks raised interest rates sharply to fight inflation. Higher real rates make non-yielding gold less attractive, which can offset the inflation effect. So gold's response to inflation depends on how policymakers respond - which is hard to predict.

A fair summary

• Over many decades, gold has broadly kept pace with or beaten inflation.
• Over periods of 1 to 20 years - the periods that matter to most retirees - it has sometimes failed badly.
• Other assets also respond to inflation. Stocks have beaten inflation over most long periods (8.4% real over 1995-2025), and Treasury Inflation-Protected Securities (TIPS) are designed specifically to track it. A fee-only adviser can explain how they compare for your situation.

If inflation protection is your main reason for considering a gold IRA, it is worth knowing that gold is an unreliable inflation hedge in the short and medium term, even if it has worked over the very long term. We answer this directly in the FAQ Does gold beat inflation?

If gold is not a reliable growth asset or a reliable inflation hedge, what is its strongest case? For most researchers and planners, it is diversification: owning something that does not move in lockstep with your stocks and bonds.

Why gold can diversify

Diversification works when assets respond to different forces. Stocks depend mainly on company earnings and economic growth. Bonds depend mainly on interest rates and credit risk. Gold depends on a different mix - real rates, the dollar, central-bank reserve policy, fear, jewelry demand. Because those drivers overlap only partly, gold's price has often moved independently of stocks. Adding a modest amount of something that moves differently can smooth a portfolio's overall ride, even if that asset is volatile on its own.

Crisis behavior: often helpful, not dependable

Gold has a reputation for rising when markets panic, and in a number of episodes it has held up while stocks fell. Investors seeking a store of value outside the financial system - especially during banking scares, currency worries or geopolitical shocks - tend to add to that demand.

But three caveats are important:

1. It does not always work. In some sharp market sell-offs, gold has fallen alongside stocks at first, as investors sell whatever they can to raise cash or cover losses elsewhere. The protection, when it comes, may arrive later.
2. Gold has its own crises. The 2026 fall of about 29% in five months happened to gold itself. A "hedge" that can lose almost a third of its value in half a year does not remove risk; it swaps one kind of risk for another.
3. Price paid matters. Gold bought after a big run-up starts from a high base. If a crisis hits after gold has already rallied hard, there may be less room for it to rise further, and more room to fall when fear fades.

The diversification case in numbers - how to think about it

We are not going to quote a single "optimal" gold allocation from a study, because results depend heavily on the period, the assets and the assumptions used, and they can be presented selectively. The principle is easier to state than to measure: a small position in an asset that sometimes zigs when others zag can reduce the depth of a portfolio's worst years, at the cost of some long-run growth. A large position simply makes your portfolio's fate depend on gold - which, as the drawdown table shows, is not a stable thing to depend on.

That is why the realistic question is usually not "Is gold a good investment?" but "Is a small amount of gold a good addition to what I already own, at the price I will pay for it?" We look at how much commentators typically suggest in how much to hold, and in depth in our guide how much gold in a retirement portfolio.

Gold is a global, liquid market - but your access to it may not be

Gold trades around the clock in London, New York, Shanghai and elsewhere, and the LBMA benchmark price is published twice each trading day. The market is deep. But the speed and cost at which you can sell depends on how you hold it. In a gold IRA, selling involves your custodian, the depository and a dealer, and you sell at the dealer's bid, typically below spot. In a crisis, dealer spreads can widen. A gold ETF in a brokerage IRA can be sold in seconds during market hours. That difference matters if your reason for owning gold is to have something you can use in an emergency.

Here are the strongest arguments for holding some gold in a retirement account - each with the evidence behind it and the catch that comes with it. We have tried to state each one as fairly as its supporters would.

1. It can diversify a stock-and-bond portfolio

Evidence: gold's price is driven by different forces from stocks and bonds (real interest rates, the dollar, central-bank buying, fear demand), and over 2000-2025 it returned about 10.4% a year while the S&P 500 returned about 7.9% - a period in which a stock-heavy retiree would have benefited from owning some gold.

Caveat: diversification helps only in modest amounts, and gold's independence from stocks is not reliable in every crisis. It also costs more to hold in an IRA than other diversifiers.

2. It has held purchasing power over very long periods

Evidence: over 1995-2025, gold returned about 5.4% a year after inflation (Kiplinger); over 1971-2025 about 8.8% a year nominal.

Caveat: over shorter periods - including the entire 1980s and 1990s and the high-inflation years of 2021-2022 - it did not. See the inflation section.

3. It is not anyone's promise to pay

Evidence: a bar of gold does not depend on a company staying in business or a government paying its debts. It has no default risk in the way a bond does. That is a big part of why central banks hold it, and why recent central-bank buying has been described as at record levels.

Caveat: "no default risk" is not "no price risk." Gold's price can still fall 29%-70%, as history shows. And in a gold IRA you rely on a custodian and a depository to hold and record your metal, so you are not free of counterparty relationships.

4. Tax advantages of an IRA

Evidence: inside a traditional IRA, gains grow tax-deferred; inside a Roth IRA, qualified withdrawals are tax-free (see our Roth gold IRA guide). Outside an IRA, gains on physical gold - and on grantor-trust gold ETFs such as GLD - are taxed as "collectibles" at up to 28% long-term, higher than the usual long-term capital gains rates.

Caveat: you can get exactly the same tax shelter by holding a gold ETF inside an ordinary IRA at a brokerage, usually at a much lower cost. The tax advantage belongs to the IRA, not to physical metal.

5. You own specific physical metal

Evidence: a gold IRA holds actual coins or bars in an approved depository, held for your account. Some people value owning a physical asset rather than a share of a fund, and segregated storage lets you hold identifiable items.

Caveat: you cannot keep IRA gold at home - doing so through an "IRA LLC" was treated as a taxable distribution in McNulty v. Commissioner (2021). So in practice you never touch the metal while it is in the IRA, which makes the "physical" benefit more psychological than practical. Our gold IRA rules guide explains storage requirements.

6. It can be moved from an existing account without tax

Evidence: a direct trustee-to-trustee transfer or direct rollover from a 401(k), 403(b), TSP or existing IRA is not taxed and has no annual limit. You do not need new cash to start.

Caveat: rolling over retirement money is exactly how many retirees have been drawn into high-markup sales. Regulators have documented markups of up to about 130% in enforcement cases. See our rollover guide before moving anything.

7. Recent momentum and strong demand

Evidence: gold hit a record annual average of $3,431 in 2025 with 53 record LBMA PM prices, and total demand reached a record 1,231 tonnes in Q1 2026.

Caveat: momentum is the weakest argument on this list. Gold fell about 29% within five months of its record. Strong recent performance is a reason prices are high, not a reason they will keep rising.

Now the other side, in the same detail. These are the drawbacks we think every buyer should understand before signing anything.

1. No income

Gold pays no interest, dividends or rent. A retiree who holds $100,000 of gold gets nothing from it until they sell some. Compare a stock fund or bond fund, which pays cash you can spend without selling. Over decades, those reinvested payments are a major part of why the S&P 500 has outgrown gold (about 11.2% vs 8.8% a year since 1971).

2. Volatility that does not suit "safe money"

Gold's yearly moves in the last decade ranged from about -10.7% to about +65%, and in 2026 it fell about 29% in five months. If you are buying gold for peace of mind, you should be comfortable seeing your gold IRA statement drop by a quarter or more without selling in panic.

3. Long flat or losing periods

Gold's history includes a 70% nominal slide from 1980 to 1999 and roughly a decade (2011-2020/2023) below its 2011 high. Over 2020-2022, the annual average price barely moved ($1,770, $1,799, $1,800). Long periods of going nowhere are a real cost when you are also paying yearly fees.

4. Fee drag

A gold IRA is among the most expensive ways to own gold. You pay:

• A dealer markup over the spot price - roughly 2-7% on common bullion at online dealers, often more at phone-sales gold IRA companies, and far more on "premium" or proof coins.
• A spread when you sell - dealers typically bid 1-2% under spot for common coins.
• Yearly custodian and storage fees - commonly about $200-$700 or more combined, depending on the custodian, account size and storage type.

On a small account, those flat fees can take 1-2% a year - more than many gold ETFs cost in five or ten years. We quantify this in the next section and in our full gold IRA fees guide.

5. Liquidity and RMD friction

Selling metal in an IRA takes several steps: you ask the custodian, the custodian arranges the sale through a dealer, the metal leaves the depository, and the cash settles. That can take days and costs a fee at many custodians (for example, Equity Trust charges a $10-per-asset liquidation fee, up to $30). Taking RMDs "in kind" means paying for shipping and handling and paying tax on the metal's value with money from elsewhere. Neither is a dealbreaker, but both are slower and costlier than selling a fund.

6. Sales pressure and scams

The gold IRA industry is heavily marketed to people aged 50 and over, often using fear of a market crash, inflation or bank failures. Enforcement cases show what can go wrong:

None of this means every gold IRA company behaves this way. It means the risk of overpaying is real and you must check prices yourself. Read our gold IRA scams and red flags guide and the joint CFTC/FINRA/NASAA advisory, 10 Things to Ask Before Buying Physical Gold, Silver, or Other Metals.

7. Opportunity cost

Every dollar in gold is a dollar not in something else. Over 1995-2025, the gap between gold's 5.4% real return and stocks' 8.4% real return turned $10,000 into roughly $48,000 vs $112,000 of 1995 purchasing power. If a large gold position causes you to miss even part of that kind of growth, the cost may be bigger than any fee.

8. Marketing that overstates safety

Watch for language like "crash-proof," "guaranteed to protect," "the IRS-approved way" or "the government may seize your 401(k)." No gold IRA carries a government guarantee of value, dealers are not "IRS-approved," and gold's own history disproves "crash-proof." Sales scripts built on fear are a reason to slow down, not speed up.

Pros and cons at a glance

Fees are the part of the gold IRA decision you can actually control, so it is worth seeing exactly how much they matter. The following examples are hypothetical illustrations using round numbers, not quotes from any company.

The setup

Two investors each move $100,000 into a gold IRA. They buy the same kind of bullion. The only differences are what they pay:

For context: $260 a year is in line with a low-cost custodian plus commingled storage (for example, STRATA's reported $150 metals tier plus $115 storage = $265). $500 is the Equity Trust annual fee for a $100,000-$199,999 account before storage ($110-$160 more). Real fees vary widely - use our gold IRA fee calculator with your own quotes.

Step 1: how much gold you actually get

• A: $100,000 / 1.05 = $95,238 of gold at spot value. The other $4,762 is the dealer's markup.
• B: $100,000 / 1.15 = $86,957 of gold at spot value. The other $13,043 is markup.

Step 2: what you would get if you sold the next day

• A: $95,238 x 0.99 = $94,286, a loss of about $5,714 (5.7%).
• B: $86,957 x 0.99 = $86,087, a loss of about $13,913 (13.9%).

That is the "hole" each investor starts in on day one, even if gold's price does not move at all.

Step 3: how much gold must rise just to break even

Break-even rise = (money put in + fees paid) / day-one sale value - 1. We assume fees are paid from outside the IRA, so they add to what you have spent.

Example of the math for B over 10 years: ($100,000 + $5,000) / $86,087 = 1.2197, so gold must rise about 22% just to get back to even. Spread over 10 years, that is 1.2197 to the power of 1/10 = about 1.020, or about 2.0% a year. Investor B needs gold to rise more than twice as fast as investor A simply to stand still.

Step 4: the same gold return, different results

Now suppose gold rises a steady 5% a year for 10 years (a hypothetical rate, not a forecast). At the end, each investor sells at 1% under spot and we subtract the fees paid along the way:

Same gold, same decade. Investor B ends up about $15,800 behind investor A and about $26,000 behind a low-cost ETF holder. (ETF figures ignore the small brokerage bid-ask spread, and the IRA figures assume fees are paid from outside the account; if fees were deducted by selling metal, results would be slightly lower again.)

Step 5: fees as a percentage depend on account size

Because custodian and storage fees are mostly flat dollar amounts, they hit small accounts hardest:

The takeaway: the smaller your gold IRA, the harder it is to justify the flat fees, and the higher the markup, the longer gold must rise before you make anything. A 15% markup is close to the point where many commentators suggest pushing back; anything above that on ordinary bullion deserves a hard look (this is a rule of thumb, not an industry standard). For a full breakdown of every fee, custodian schedules and how to negotiate, read our gold IRA fees guide.

If you decide you want some gold exposure, a gold IRA is only one of four common routes. Each is better for a different purpose.

Gold ETF expense ratios

Expense ratios (the yearly fund fee as a percentage of your holding) for well-known funds, as of our latest check - verify before you buy:

When each route tends to be better

• Gold ETF in an existing IRA is usually the cheapest and simplest way to add a small gold slice to retirement savings. No new custodian, no storage bill, no dealer markup, and you can rebalance in seconds. It suits most people whose goal is diversification. Our gold IRA vs gold ETF comparison goes deeper.
• A gold IRA can make more sense if you specifically want identifiable physical bullion held for you, have a large enough amount (so flat fees are a small percentage), and can buy at a competitive markup. Some people also simply distrust fund structures.
• Coins at home suit people who want metal they can hold, outside the financial system, and who accept theft risk and the collectibles tax rate. It cannot be done with IRA money.
• Mining stocks are a bet on companies, not on gold. They often rise more than gold in good years and fall more in bad ones, and they carry business risks (costs, management, politics). They are a different investment, not a substitute.

Put simply: if your goal is "some gold exposure at the lowest cost," a gold ETF usually wins. If your goal is "own specific physical bullion in a tax-advantaged account," a gold IRA is the main option - and the job is to keep its costs as low as possible.

There is no single answer, but some situations fit a gold IRA better than others. These profiles are general illustrations, not recommendations.

Signs a gold IRA is probably not right for you now

• You would need to sell the gold within the next few years.
• You are relying on the account for monthly income.
• You are considering it mainly because of a phone call, advertisement or "free silver" offer.
• You would be putting in a large share of your total savings.
• You cannot get written prices showing the markup over spot.
• Your account is small enough that the yearly fees exceed about 1% of it.

Signs it may be worth exploring

• You already have a diversified plan, with emergency cash and several years of withdrawals in stable assets.
• You want a small physical-metal slice and understand it may fall 30% or more.
• Your account is large enough that flat fees are a small percentage.
• You are willing to compare prices from several dealers and walk away from pressure.
• You have discussed it with a fee-only adviser who has no commission on the sale.

There is no official answer, and nobody can tell you the right amount without knowing your full situation. As general education: commentators and planners who include gold commonly cite roughly 5-10% of a portfolio as a typical range, and some suggest less or none. You will hear much higher figures from companies that sell gold - remember who is giving the advice.

What those percentages mean in dollars:

Notice the bottom of the first row: $10,000 is the minimum at some gold IRA companies, and at that size $260-$500 a year in fees would be 2.6%-5% of the gold. That is why, for smaller portfolios, a low-cost gold ETF inside an existing IRA is often a more practical way to hold a 5-10% slice.

Rebalancing: a worked example

Choosing a percentage is only half of it. The other half is rebalancing - periodically trimming whatever has grown too big and adding to what has shrunk, to get back to your target. This is a hypothetical illustration:

1. You have $500,000 and target 10% in gold: $50,000 gold, $450,000 other.
2. Gold rises 65% in a year (as it roughly did in 2025): $50,000 x 1.65 = $82,500. For simplicity, assume everything else stays at $450,000.
3. Your total is now $532,500, and gold is $82,500 / $532,500 = 15.5% - well above your 10% target.
4. 10% of $532,500 is $53,250. To rebalance, you would sell $82,500 - $53,250 = $29,250 of gold and move it into your other holdings.

Rebalancing forces you to sell some of what has risen and buy what has lagged - in this case, it would have locked in part of the 2025 gain before the 2026 drop. It works the other way too: after a fall, you would add to gold. In a gold IRA, rebalancing involves dealer spreads and possibly custodian fees each time, which is one more reason some people hold their gold slice in a low-cost ETF instead.

For a fuller discussion of allocation frameworks, see how much gold in a retirement portfolio. Our FAQ What percentage of my retirement should be in gold? gives the short version.

Many people start thinking about gold after it has already made headlines - which, in 2026, means after it rose from about $2,063 at the end of 2023 to about $5,595 in January 2026 and then fell back to about $4,170. Is it too late? Nobody knows. What you can control is how you buy.

Two approaches

• Lump sum: invest the whole amount at once. If the price then rises, you benefit fully; if it falls, you take the full hit.
• Dollar-cost averaging (DCA): invest equal amounts at regular intervals - for example, monthly over six months. You buy more ounces when prices are low and fewer when they are high, and you avoid putting everything in at a single unlucky moment.

Two hypothetical examples

You have $60,000 to put into gold. Compare buying all at once with $10,000 a month for six months. The prices below are invented for illustration, not forecasts.

Scenario 1 - prices fall, then partly recover: monthly prices of $5,000, $4,600, $4,200, $4,000, $4,200, $4,400.

• Lump sum at $5,000: $60,000 / $5,000 = 12.0 oz, worth $52,800 at $4,400.
• DCA: $10,000 / each price, added up = about 13.7 oz, an average cost of about $4,377, worth about $60,300 at $4,400.

Scenario 2 - prices rise steadily: $4,200, $4,400, $4,600, $4,800, $5,000, $5,200.

• Lump sum at $4,200: $60,000 / $4,200 = about 14.3 oz, worth about $74,300 at $5,200.
• DCA: about 12.8 oz, an average cost of about $4,675, worth about $66,700 at $5,200.

So DCA did better when prices fell and worse when they rose. That is the general pattern: because markets more often rise than fall over time, a lump sum has tended to come out ahead more often than not, but DCA reduces the regret and damage of buying right before a drop. For a volatile asset bought after a big run, many people value that protection.

DCA inside a gold IRA has extra costs

Each purchase in a gold IRA may carry a custodian transaction fee (for example, STRATA reportedly charges $40 a trade plus depository handling) and each dealer order has its own markup. Six small purchases can cost more than one large one. Ask your custodian about per-trade fees before planning a DCA schedule - or consider averaging into a low-cost ETF instead, where small regular purchases are cheap.

What not to do

• Do not rush because a salesperson says prices are "about to explode" or that an offer ends today.
• Do not move more than you planned because the price just dropped and "it's on sale." A falling price can keep falling.
• Do not use money you will need for living costs in the next few years.

Precious metals IRAs can also hold silver (.999 fine), platinum (.9995) and palladium (.9995). They are often pitched as cheaper or "higher upside" alternatives to gold. Briefly:

Silver

• Annual average prices: 2015 $15.68 - 2016 $17.14 - 2017 $17.05 - 2018 $15.71 - 2019 $16.21 - 2020 $20.55 - 2021 $25.14 - 2022 $21.73 - 2023 $23.35 - 2024 $28.27 - 2025 $40.03.
• 2025: closed around $72, up about 145% for the year.
• Record: reportedly about $121.62 on January 29, 2026 (one source; not independently verified).
• Now: about $57.70 (October 19, 2026) - roughly 53% below that reported record ($57.70 / $121.62 - 1 = -0.526).
• Gold/silver ratio: about 72:1 ($4,170 / $57.70 = 72.3), meaning one ounce of gold buys about 72 ounces of silver.

Silver's swings are larger than gold's - a 145% gain followed by a roughly 50% fall within about a year. It also has heavier industrial demand, so it is more sensitive to the economy. And it typically costs more to buy: American Silver Eagles averaged about a 6.87% premium in August 2026, versus about 3-6% for Gold Eagles, and silver takes far more storage space per dollar. See our silver IRA guide.

Platinum and palladium

Both are much smaller markets, driven heavily by industrial use (especially vehicle catalytic converters). That makes them more volatile and more tied to manufacturing cycles than gold. Premiums can be high: Platinum Eagles averaged about 9.21% over spot in July 2026 and about 13.35% in February 2026. For most retirement savers, these metals are niche holdings at best.

The bottom line on other metals

If gold is too volatile for a given role in your plan, silver, platinum and palladium are more so. They can add diversification, but they amplify the downsides discussed in this guide - and a salesperson steering you toward them, or toward "premium" coins of any metal, may be steering you toward higher markups.

Before you contact any gold company, work through these questions on paper. If you cannot answer one, that is a sign to pause. They are also good questions to bring to a fee-only adviser.

About your goals

1. What job do I want gold to do? Diversification, inflation worry, crisis protection, or "it's been going up"? Be honest - the last is the weakest reason.
2. What would I hold instead? Name the money's alternative use, and compare.
3. When will I need this money? If within about five years, gold's 29%-70% historical drops are a serious risk.

About the amount

4. What share of my total savings would this be? Commentators commonly cite 5-10%; are you above that, and why?
5. Could I live with a 30% fall? On a $50,000 gold position, that is a $15,000 paper loss. On $200,000, it is $60,000. Would I sell in panic?
6. Do I have emergency cash and near-term withdrawals covered elsewhere?

About the costs

7. What is the markup over spot, in writing, for the exact products? If a company will not say, walk away.
8. What will they pay me if I sell back tomorrow? The gap between the two prices is your real round-trip cost.
9. What are the yearly custodian and storage fees, and what percentage of my account are they? Run them through our fee calculator.
10. Could a gold ETF in my existing IRA do the same job for less?

About the seller

11. Is anyone pushing me toward proof, "premium" or "collectible" coins? These usually carry much higher markups.
12. Am I being rushed, frightened or offered "free" metal to decide quickly?
13. Have I checked the company with regulators and complaint databases, and read the CFTC/FINRA/NASAA advisory?

About taxes and the rules

14. Will I move the money by direct transfer or direct rollover (not a 60-day rollover that risks taxes and the 20% withholding from employer plans)?
15. How will I take RMDs - by selling metal or in kind - and have I asked a tax professional about the effect?
16. Do I understand that the metal cannot be stored at home?

If your answers point to a small allocation, low costs, money you will not need soon and no pressure, a gold IRA may be a reasonable choice for you. If they point the other way, there is no shame in deciding against it - or in choosing a cheaper route to the same exposure.

This section is for readers who have weighed the evidence and decided a gold IRA fits their plan. It includes links to our company reviews. Gold IRA Planet may receive compensation from some companies we review; this does not change our assessment of the evidence above.

1. Learn the process

• Read our step-by-step gold IRA rollover guide to move money from a 401(k), TSP or IRA without triggering taxes.
• Read the gold IRA rules on eligible metals and storage, and the taxes and RMDs guide.
• If you are weighing a Roth account, see the Roth gold IRA guide.

2. Choose a company carefully

• Use our guide on how to choose a gold IRA company - it lists the questions to ask and the answers that should worry you.
• See our gold IRA company rankings and how we rank companies.
• Get written quotes from at least two or three companies for the same products, and compare markups over spot.

3. Company reviews

Our reviews of three widely advertised companies (figures reported by third parties; confirm in writing):

None of these companies publishes its product markups. Whatever company you consider, ask for the spread in writing before you fund the account, and remember the break-even math: a lower markup matters more than any promotional "free silver."

4. Keep it in proportion

Decide your target percentage before you talk to anyone, write it down, and do not let a sales call move it. Review it once a year and rebalance if gold has drifted well above or below your target.

No one can say whether gold will rise or fall from here. As of October 19, 2026, it is about $4,170 an ounce - about 25% below its January 29, 2026 record of about $5,595, about 4% above its June low of about $3,994, and still about 59% above where it ended 2024. Bank forecasts we have seen for 2026 range from about $4,325 to $5,000, and forecasts made before January did not predict the spike or the drop. Gold's case in 2026 is the same as in any year: a possible diversifier in modest amounts, bought at low cost. Its 2024-2025 gains are not a reason to expect more, and its 2026 drop is not a promise of a bargain.

Yes. You can lose money in three ways. First, the gold price can fall: it dropped about 29% in the first half of 2026, about 45% from 2011 to 2015 and about 70% from 1980 to 1999. Second, costs create an immediate loss: with a 5% markup and a 1% sell spread, you would get back only about $94,300 of a $100,000 purchase if you sold the next day, and with a 15% markup only about $86,100 (see our worked examples). Third, yearly fees continue whether gold rises or not. There is no government insurance on the value of the metal. Fraud is a further risk: regulators have documented markups of up to about 130%.

Over very long periods, broadly yes: over 1995-2025, gold returned about 5.4% a year after inflation (about 8.1% before). But over periods that matter to most retirees, it has often failed. From 1980 to 1999, gold fell about 70% in dollar terms while the cost of living kept rising. During the 2021-2022 inflation surge, gold's reported yearly change was about -5.8% and +0.4%, and its average price was almost unchanged ($1,799 to $1,800). Gold is better described as a long-run store of value with unreliable short-run inflation protection. See our inflation section for detail.

They are not really alternatives. A 401(k) is an employer plan, often with a company match - which is free money that a gold IRA cannot replace - and usually low-cost funds. A gold IRA is a self-directed IRA that holds physical metal. Many people keep their 401(k) for growth and, if they want gold, hold a small slice either through a gold fund (if their plan offers one), a gold ETF in an IRA, or a gold IRA funded by a partial rollover. Over the long run, US stocks have outgrown gold (about 11.2% vs 8.8% a year since 1971), though not in every period. Before rolling a 401(k) into gold, read our rollover guide and consider whether you would lose a match or lower-cost funds.

There is no rule. As general education, commentators who include gold commonly cite roughly 5-10% of a portfolio; some suggest less or none. Above that, your retirement increasingly depends on one volatile asset that pays no income. The right figure depends on your age, other income, how soon you need the money and how you would feel about a 30% drop. Very high percentages tend to come from people selling gold. A fee-only adviser can help you set a target and a rebalancing plan. Our guide how much gold in a retirement portfolio explains the trade-offs.

Nobody knows. Gold rose about 65% in 2025 and reached a record in January 2026, then fell about 29% by June. Someone who bought at the end of 2025 was about 3% behind on October 19, 2026 on spot prices alone - more after costs - while someone who bought at the end of 2023 was about 100% ahead. History shows gold can keep rising after a big run, or go sideways or down for years. If you do want a gold slice, buying gradually (dollar-cost averaging), keeping the amount modest and minimizing costs reduces the damage of bad timing. See lump sum vs dollar-cost averaging.

Here is our honest summary of 50 years of data:

• Gold has been a real investment, not a "pet rock." Since 1971 it has returned about 8.8% a year, and about 5.4% a year after inflation over 1995-2025. In some long periods, such as 2000-2025, it beat stocks.
• But it has usually grown more slowly than stocks (about 11.2% a year since 1971 for the S&P 500 with dividends), and it pays no income.
• It is volatile. Drops of about 29% (2026), 45% (2011-2015) and 70% (1980-1999) - with recoveries of a decade or more - are part of its history, and matter more the older you are.
• It is an unreliable short-term inflation hedge, as the 1980s-90s and 2021-2022 showed.
• Its best case is diversification in modest amounts, commonly 5-10% of a portfolio in commentators' examples.
• A gold IRA is the costliest way to hold it. A 15% markup and $500 a year means gold must rise about 22% over 10 years just to break even; a gold ETF in your existing IRA can cost 0.09%-0.40% a year.

So, is a gold IRA a good investment? For someone with a solid, diversified plan who wants a small slice of physical bullion, can get it at a low markup and does not need the money for years, it can be a reasonable choice. For someone moving a large share of savings out of fear, buying expensive coins on a sales call, or holding a small account where fees take 1-2% a year, it is likely a poor one.

Whatever you decide, decide slowly. Get prices in writing, compare at least two or three sellers, read the scam red flags, and talk to a fee-only financial adviser who earns nothing from the sale. If you are still learning the basics, our complete gold IRA guide is the place to start.

This article is general information, not financial, tax or legal advice. Fees, minimums and promotions change often and vary by custodian and account size - confirm current terms in writing with the company and your custodian before you invest.