Gold After Selling a Business: The Asset With No Cash Flow

Gold after selling a business is a form of insurance, not a replacement for the company that produced the proceeds. It pays nothing, employs no one, and can fund a life or a tax bill only by being sold. Any return comes entirely from a later buyer paying more. That makes it a different instrument from an operating business, and it has to be judged by a different test.

In May 1999, Britain's Treasury, under Chancellor Gordon Brown, told the market it intended to sell a little over half of the nation's gold reserves.¹ An ounce that had changed hands for roughly $850 at the start of 1980¹ was then trading near $280.¹ The metal paid no interest and cost money to store. Several other central banks — Belgium, Canada, the Netherlands among them — had already been lightening their holdings through the decade.² The stated aim was to swap a volatile, idle reserve for interest-bearing assets in dollars, euros and yen. It was a treasurer's instinct, not a trader's.

The programme ran through seventeen pre-announced auctions between 1999 and 2002.³ Roughly 395 tonnes left the vaults³ at an average of around $275 an ounce,¹ raising in the region of $3.5 billion,³ which went straight into foreign government paper. The announcement itself moved the price: traders now knew a large, patient, publicly scheduled seller was coming. Gold touched a low near $253 that July.¹ Within a few years the metal was rising, and it kept rising for most of a decade. London dealing rooms gave the trough a nickname that has never quite faded — Brown's Bottom.¹

The episode is almost always retold as a lesson about timing. That is the least useful thing about it. The mechanics of the sale were not the scandal; the auctions were transparent and orderly, and the reasoning behind them was, in one important respect, entirely correct. The decision was still poor. Understanding how both of those can be true at once is the whole of the lesson for owners who have just converted a company into a bank balance and are looking at elevated metal prices and stirring inflation numbers, wondering whether to buy some protection.

What the Treasury got right before it got the price wrong

The official case rested on two observations. Gold was an uncomfortably large and jumpy share of Britain's reserves — a concentration problem. And gold pays no interest, while bonds do. A reserve held in government paper earns something every year. A reserve held in bullion sits in a vault beneath the City of London, costing money to store, insure and audit. Swapping idle metal for income-producing paper looked like plain good housekeeping.

Both observations were true. Gold really does produce nothing, and that is not a temporary condition to be waited out — it is what gold is. An ounce held for fifty years is the same ounce at the end: no interest paid, no dividend declared, no earnings retained and reinvested. In a well-known illustration, Warren Buffett observed that all the gold ever mined would form a single cube small enough to sit comfortably inside a baseball infield,⁴ and that a century later it would still be a cube of exactly that size, having done absolutely nothing in the interim.

So the premise was sound. What failed was the question the premise was asked to answer. The Treasury asked, in effect, which asset earns more? Bonds win that contest against gold every time, by definition, because one pays a coupon and the other pays nothing. But a national reserve is not held in order to earn. It is held to be there, in a universally accepted form, in precisely those circumstances when paper promises come under strain. The Treasury judged an insurance policy by its running yield, found the yield to be zero, and cancelled the policy — at the end of a long calm, when insurance looks most pointless and is therefore cheapest.

It was right about the yield and wrong about the job. That distinction — between what an asset earns and what job it has been hired to do — is the portable part of the story. Everything else about gold after an exit follows from it.

The bid is the only thing that moves

It is worth being precise about where gold's return comes from, because gold is unusual in having exactly one source.

Owners who have run a company know in their bones that a business generates cash, and that the cash arrives in the bank whether or not anyone outside has an opinion about the firm that quarter. Over long enough periods the price someone will pay for a business is dragged along by that cash. That is the argument an entire section of this library already makes at length, and it needs no re-arguing here.

Gold has no such anchor underneath it. There is no cash to pull the price anywhere over time. The entire return from holding it — every dollar of it — is the gap between what one buyer paid and what a later buyer agreed to pay. Jewellery and industrial demand exist, but they do not turn a bar in a vault into a business. If nobody's view changes, nothing happens. The metal does not grow, enter new markets, raise its prices, cut a cost or buy back its own shares. Its price is a running tally of what other people feel: how much they trust the currencies they hold, how nervous they are about banks and governments, how many central banks decide to add to their vaults in a given year.

That is a description, not a criticism. But it has two consequences that matter enormously to someone whose expertise was built inside an operating company.

The first is that a founder's judgement has no purchase on it. Someone who spent twenty years in manufacturing can look at a company and tell whether customers will keep coming back, whether suppliers have the firm over a barrel, whether margins are real or borrowed. That instinct — the habit of learning a business from the people around it — is genuinely transferable to owning shares in other companies. It is worthless against a cube of metal. There is nothing to ask, no customer to call, no supplier with a view. The only question that determines gold's price is what a large and shifting crowd will feel about money in five or fifteen years. That is a forecast about collective psychology, and business owners have no particular edge in making it.

The second consequence is less comfortable, and the buyers of 1980 illustrate it better than the sellers of 1999 do.

The twenty years the last panic's buyers spent going backwards

In January 1980, with inflation running in double digits,⁹ oil shocks fresh in memory and geopolitics leading the evening news, gold touched roughly $850 an ounce.⁵ Those purchases were insurance too. They were made by people who had watched paper money lose its purchasing power right through the 1970s and had no intention of being paid in a shrinking unit. Their reasoning was not foolish; it was the same reasoning that circulates whenever wholesale prices reaccelerate.

Then, under Paul Volcker, the Federal Reserve raised interest rates high enough to break the inflation⁸ and to make cash and bonds handsomely competitive again. The fear that had bid the metal up receded. From that peak, gold fell in fits and starts for the better part of two decades — on the order of seventy percent in nominal terms — until it was trading in the mid-$200s around the time the British auctions began.⁶ Ordinary prices did not fall during those twenty years. They roughly doubled.⁷ Anyone who bought gold at the 1980 peak as protection against rising prices spent a generation going backwards in money terms, and considerably further in purchasing power, precisely while the thing they were hedging against carried on happening.

The mechanism explains it. Gold does not respond to the price index. It responds to fear about the price index, and to fear more generally — and fear does not move in step with the cost of groceries. It moves in surges and long lulls. The 1980 peak was the culmination of a decade of anxiety about currencies; what followed was twenty years of growing confidence in central banks, during which the metal drifted down while inflation quietly continued. The hedge worked when the crowd was frightened and failed when the crowd was calm, largely regardless of what prices were actually doing.

There is a plainer way to say the same thing. The real competitor to gold is not the stock market; it is the interest available on safe money after inflation is subtracted — what economists call the real rate. When high-quality bonds pay more than prices are rising, holding something that pays nothing is a decision to earn less on purpose, and the crowd tends to notice. When that spread compresses or turns negative, the penalty shrinks. Nothing has changed inside the metal either way. Only the price of refusing a coupon has changed.

This is why gold is a poor instrument against ordinary, year-by-year inflation, and why it is a fundamentally different thing from owning a business that can pass higher costs through to its customers whether or not anybody is nervous — a distinction worked through in detail elsewhere in this library. One asset adapts. The other waits.

Waiting has a cost that is easy to underestimate on a twenty- or thirty-year family horizon. Twenty years is long enough for children to grow up, for a second business to be built and sold, for an estate to change hands. A holding that must be liquidated at whatever the bid happens to be in order to fund any of that is not really performing the function it was bought for. A hedge that has to be sold to live is a hedge that is not in place.

What gold is actually insuring against

None of this makes gold useless, and the British episode is the proof. The Bank of England had held bullion for centuries not because it earned anything but because there are states of the world in which almost nothing else is trusted.

The risk in question is one founders rarely think about while running a company and cannot stop thinking about the week after the sale closes: counterparty risk — the fact that nearly every financial claim is simultaneously somebody else's obligation. A bank deposit is a promise from a bank. A government bond is a promise from a government. A share is a claim on a company's future, settled through clearing houses and custodians. While a founder ran the business, the wealth was a factory, a customer list and a payroll. The morning after the wire lands, it is a ledger entry in a system of interlocking promises.

Physical gold is not a promise from anyone. It does not require a solvent custodian, a functioning settlement system or a cooperative government in order to exist. That is the narrow, genuine job it does: it insures against a breakdown of confidence in the paper system itself, not against next year's cost of living. And the right way to judge an insurance policy is never by what it earns in the years nothing happens. It is by whether it pays out on the day something does.

Which is exactly why the sellers of 1999 lost the argument. Not because gold "performed" — it never performs; it has no operations — but because the decade that followed brought a financial crisis, emergency intervention on a scale not seen in generations, and a broad wave of public doubt about the soundness of money. The Treasury had traded an asset that pays out when confidence collapses for assets whose value depends on confidence being maintained. The policy it cancelled would have paid.

The form the insurance takes changes what is being insured, and this is where a good deal of confusion lives. The comparison below is about structures, not outcomes.

The trade-off is symmetrical and there is no free version. The paper form removes the logistics and reintroduces the institutions. The physical form removes the institutions and reintroduces the logistics — transport, vaulting, audit, insurance, all of which cost money every year. Either way, a holding that generates no income carries a small negative yield before anyone's view of the world has changed at all. Families who hold gold deliberately tend to treat that cost as what it is: a premium, not an investment expense. Which is also, incidentally, why insurance is sized the way it is — small enough that years of paying for nothing are tolerable, large enough that a payout would actually matter. A family that puts half its proceeds into metal has not bought insurance. It has made a large, undiversified wager on the mood of strangers.

Naming the job before buying the instrument

The useful question after an exit is not whether gold will go up. Nobody can answer that reliably, and this publication does not attempt price forecasts. The useful question is the one the Treasury skipped: what job does this money have, and is this the right instrument for that job?

Proceeds from a sale have several jobs at once, and they are not interchangeable. Some of the money has to pay a tax bill on a known date. Some has to fund a family's spending for decades, which means it has to produce something rather than merely be worth something — living off cash a business generates is a genuinely different problem from living off a price. Some may be earmarked for a next venture. And some, for certain families, is set aside against the possibility that the ordinary financial arrangements everyone relies on stop working — capital they would be perfectly content to watch sit flat for twenty years if the crisis never arrives.

Gold can only ever serve the last of those. Hired for any of the others, it will fail, and it will fail slowly enough that the failure is easy to miss until the moment it matters.

That is the habit worth carrying out of this story, and it extends well past one metal. Every instrument put in front of a newly liquid owner — a structured note, a private fund, a building, a bar of gold — should be made to answer the same two questions in order: what job is it being hired for, and by what test will it be judged? Britain's Treasury got the second question right and never asked the first. It applied an earnings test to an asset that was never held to earn. Owners in this position have a natural advantage here, because they spent years refusing to buy machines without knowing what those machines were for. The discipline transfers. It is only the object that changed.