The gold price is the same for everyone. What you pay is not.
Buy one troy ounce of gold in five different physical forms on the same afternoon and you will pay five different prices, sometimes differing by several percent for exactly the same quantity of metal. That gap is the premium over spot, and it is the single largest determinant of what a physical gold purchase actually returns.
This article explains where the premium goes, when paying a larger one is rational, when it is not, and how the market actually catches counterfeits. The last part matters because the counterfeits that circulate now are considerably better than the ones people describe in comment sections.
A retail gold product's price above spot is not one number. It is a stack.
Refining and fabrication. Turning Good Delivery bars into small bars or coin blanks costs real money in energy, labour and equipment, and that cost is per item rather than per ounce.
Minting and design. Sovereign coins carry engraving, security features, quality control and often legal tender status. Some programmes add capsules, certificates or assay cards.
Distribution. Mints sell to authorised distributors, who sell to dealers, who sell to you. Each layer covers costs and takes a margin.
Inventory and price risk. A dealer holding stock is exposed to the gold price between acquiring and selling, and hedging that exposure costs money.
Dealer margin. The actual profit, which is usually the smallest layer in the stack and the one everybody assumes is the largest.
The structural consequence is the important one: the smaller the unit, the higher the percentage premium, because the per-item costs do not shrink proportionally with the metal.
That last row deserves a separate warning. A collectible coin's value depends on grading, rarity and collector demand, which are skills and markets entirely separate from gold. Buying numismatics as a way to own gold is buying two risks and understanding one.
The case for coins. Sovereign bullion coins from established mints are the most recognisable gold objects in the world. A dealer anywhere will know what a Krugerrand, a Maple Leaf, a Britannia or an Eagle is, and will price it without needing to assay it. That recognition is liquidity, and liquidity is worth paying for. Small denominations also let you sell part of a holding, which a large bar does not. In some jurisdictions, legal tender coins receive specific tax treatment, and in the UK certain sovereign coins are exempt from capital gains tax for individuals, which can exceed the entire premium in value. Check the position in your own jurisdiction rather than assuming.
The case for bars. You get more metal per unit of currency. If your goal is the maximum gold exposure per pound or dollar spent, and you are storing it professionally rather than selling piecemeal, bars win on cost and the margin is not small over a large holding.
The case against both, for some people. Physical gold in any form carries storage, insurance and verification costs that recur every year, plus a spread you pay twice. For a holder who wants price exposure and has no requirement for possession, that cost stack is pure drag. We total it up in What It Really Costs to Own Physical Gold, and the alternative wrappers are compared in Digital Gold Investment.
Buyers fixate on the premium when they should be measuring the round trip.
The question is not "what do I pay above spot?" It is "if I bought this today and sold it back tomorrow, how much would I lose?"
That figure is the bid-ask spread on the specific product, and it has two components: what you pay above spot when buying, and what you are offered below spot when selling.
Two products with the same buy premium can have very different spreads. A widely recognised sovereign coin typically buys back close to spot or above it. An obscure private mint bar can buy back at a discount, because the dealer will need to assay or refine it. This is the practical reason recognition is worth paying for, and it only becomes visible at the exit.
How to check before buying: ask the dealer what they will pay you for the identical item today. Reputable dealers publish buy prices alongside sell prices. If a seller will not quote a buy-back price on their own product, you have learned something important.
The folk wisdom on fake gold is out of date. Gold-plated tungsten is the serious threat, because tungsten's density is very close to gold's, which means a tungsten-cored bar can pass both the weight check and the dimension check that most people believe are sufficient.
What actually works, roughly in order of reliability:
Provenance. Buying from accredited dealers, in original packaging, with assay cards and matching serial numbers, from refiners on the LBMA Good Delivery List. This is not a test, it is the avoidance of the problem, and it is what professionals rely on.
Ultrasonic thickness testing. Sound travels through gold and tungsten at very different speeds. An ultrasonic tester detects a core discontinuity that weight and dimensions cannot. This is the standard tool for larger bars.
Electrical conductivity testing. Devices measuring conductivity through the object identify internal composition differences. Effective and non-destructive.
X-ray fluorescence. Precise on surface composition and used widely in the trade. Its limitation is that it reads the surface, so it detects plating but not necessarily a deep core.
Specific gravity. Useful, cheap, and beaten by tungsten. Worth doing and not sufficient alone.
The magnet slide and the ping test. Gold is not magnetic and has a characteristic ring, so these catch crude fakes. They are a first filter, not an answer.
Practical rules that prevent most problems: buy from accredited dealers, keep the assay packaging sealed, prefer widely recognised products, treat a below-market price as a warning rather than a bargain, and have anything acquired privately professionally verified before you rely on it.
The vocabulary for reading the markings on a legitimate bar is in Troy Ounces and Gold Purity Explained.
Every form of gold ownership has an authentication question. Physical holders answer it with assay and provenance. Other wrappers answer it differently, and it is worth seeing the parallel.
ETF holders rely on a published bar list and an independent audit of the trust's allocated holdings. You are not verifying metal, you are verifying a document produced by a party with a reputation at stake.
Tokenized gold holders face a version of the same problem with a different toolkit. The metal is verified through published reserves, independent attestation and the ability to reconcile issued supply against reported backing. Some of that is checkable by anyone with a block explorer, and some depends entirely on the issuer's disclosure. The procedure is in How to Verify Gold Token Reserves Yourself.
The honest summary: physical gold lets you verify the object and requires you to trust nobody, which is exactly why people want it. It also charges you a premium, a spread, storage and insurance for that privilege, and most holders never calculate the total. Neither approach is free of trust or free of cost. They distribute both differently.
Are gold coins better than gold bars? For small holdings, partial selling and easy resale, coins are usually better because recognition is liquidity. For maximum metal per unit of currency with professional storage, bars are better. Neither dominates.
What is a normal premium over spot for gold? It varies by product size, by market conditions and by dealer, and it widens when retail demand spikes. Rather than memorising a figure, compare live quotes across several accredited dealers for the identical product on the same day.
Why is the buy price lower than the sell price? Because the dealer carries inventory, price risk and handling costs, and must make a margin on the round trip. The gap is the spread, and it is the true cost of a short holding period.
How can I tell if a gold bar is fake? Provenance first: accredited dealer, sealed assay card, matching serial number, refiner on the Good Delivery List. For testing, ultrasonic and conductivity methods detect tungsten cores that weight and dimension checks miss.
Is a tungsten-filled gold bar detectable? Yes, reliably, with ultrasonic or conductivity testing. It is not reliably detectable by weighing, measuring or the magnet test, which is why those alone are insufficient.
Should I buy small gram bars? They carry the highest percentage premiums because fabrication and packaging dominate the price. They are useful for gifting and for very small budgets. They are an expensive way to accumulate metal.
Are collectible coins a good way to own gold? They are a way to own two different things at once. The numismatic value depends on grading, rarity and collector demand, which are separate markets with separate expertise requirements. If your goal is gold exposure, bullion products are the direct route.
The premium on physical gold is mostly real cost, not mostly markup, and it rises as the units get smaller. The number worth measuring is the round trip: premium paid on the way in, discount taken on the way out, plus every year of storage and insurance in between.
If you want possession, buy recognised products from accredited dealers, keep the packaging intact, and check the buy-back price before you buy rather than after. If what you actually want is exposure to the gold price rather than custody of an object, be honest about that, because the cheapest way to hold metal in a safe is still not cheap. The full comparison of wrappers is in What Is a Gold Token.
This article is for informational purposes only and is not financial advice.
| Form | Relative premium over spot | Why |
|---|
| Good Delivery large bar | Lowest | Wholesale standard, minimal fabrication per ounce, institutional only |
| Kilobar | Low | Efficient fabrication, high metal content per item |
| 100 gram bar | Moderate | Retail accessible, reasonable efficiency |
| 1 troy ounce bar | Higher | Popular size, more fabrication cost per ounce |
| 1 troy ounce sovereign coin | Higher still | Minting, security features, brand and recognition |
| 1 to 10 gram bars | Highest | Fabrication and packaging dominate the price |
| Numismatic or collectible coins | Variable and often very high | You are buying rarity and condition, not metal |
