Does gold actually protect you?

Almost everyone arrives at gold with the same question: does it protect you when things go badly?

The short answer is that it has sometimes behaved very differently from equities during stressed periods, and that there is no mechanism which requires it to do so again. Everything useful about gold sits in the gap between those two halves of the sentence.

What gold is, structurally

Start with what is absent, because that is what makes it unusual.

Gold has no issuer. Nobody stands behind it, which means it carries no risk that a company fails or a government defaults — and also that there is nobody whose obligations you can examine.

It has no contract. A bond is an undertaking to pay stated amounts on stated dates. Gold offers nothing of the kind.

And it produces no internal cash flow. A share can pay a dividend because the business behind it earns money. A bond pays a coupon because someone is contractually required to. Gold generates nothing at all while you hold it. Storage, in fact, costs money rather than earning it.

What remains is an asset whose entire return has to come from one place: a change in its price.

The distinction people get wrong

That last fact is routinely compressed into "gold has no return", and the compression loses something important.

Producing no cash flow describes where a return would come from. It does not describe whether a return should be expected. Those are separate claims, and only the first is settled by the absence of a coupon.

The difference matters because it changes what you can argue about. Whether gold's price should be expected to rise over the long run — and by how much, relative to holding a currency or a share — is a real question, and thoughtful people give different answers to it. What you cannot do is close that question by pointing out that gold pays no dividend. That observation tells you gold's return is entirely a price return. It tells you nothing about the size of it.

The reverse error is just as common, and worth naming since it hides inside the same sentence: because gold has no cash flow to anchor a valuation to, there is no equivalent of profits or coupons against which to check whether the price is reasonable. That does not make the expected return zero. It makes it harder to estimate — which is a different and more honest complaint.

Why it sometimes behaves differently

Gold's usual role in a portfolio rests on the observation that it has often moved independently of equities, and occasionally in the opposite direction during periods of acute stress. That observation is real.

The reasoning behind it is also intelligible. Gold has no earnings to be revised down in a recession, no issuer to default, and no tie to any particular currency — so several of the things that hurt shares and bonds simultaneously do not apply to it. In periods where confidence in institutions or currencies is itself the problem, an asset that depends on none of them can be bid up precisely because everything else is being sold.

But none of that is a mechanism. There is no link in gold's construction that makes its price rise when equities fall. It has fallen alongside equities — notably in episodes where investors were selling whatever could be sold to raise cash — and it has drifted for long stretches while equities rose steadily. Describing gold as a hedge overstates the relationship; describing it as an asset driven by different things is accurate and much less comforting.

The distinction is practical, not pedantic. An asset that tends to behave differently is useful in a portfolio for reasons of composition. An asset that reliably rises when another falls would be insurance, and would be priced accordingly. Gold is the first thing and gets described as the second.

The cost, stated plainly

The counterweight to all of this is easiest to see when equities are doing well.

An asset that pays nothing has an opportunity cost, and it accumulates quietly. A share held through a flat year may still pay a dividend; a bond pays its coupon whatever the year was like. Gold held through a strong bull market in equities has produced whatever its price did, against an alternative that produced returns and income both. Over a long enough good stretch, holding gold can look like a mistake in hindsight.

This is not a defect that a better version of gold would fix. It is the same characteristic seen from the other side: an asset with no tie to corporate earnings does not participate when corporate earnings are what is driving returns.

Be careful with how far that is pushed, though, because it is the easy overstatement here. Producing no cash flow explains why gold does not participate in an earnings-driven rise. It does not by itself explain what gold's price does respond to, or why it has sometimes risen while equities fell — that depends on demand for the asset, which is a separate question with no equally tidy answer. The absence of an internal return is one reason gold behaves unlike equities. It is not the whole of the reason, and treating it as the whole is how a characteristic gets turned back into a mechanism.

What it does not do

It does not pay you. Any return is a price return, and it may be negative for years at a time.

It does not automatically offset a fall in equities. It has sometimes done so, by no rule that requires it, and expecting the pattern to hold each time is the most common way of being disappointed by it.

And it does not have a value that can be checked against anything it produces. What settles gold's price is what others are willing to pay for it — which is true of every asset in the short run, and true of gold in the long run as well.