Gold and Silver: What the Vaults Know

Anyone who has had to value an estate for probate learns early that there is no such thing as “the” price of anything. There is the price at which a thing was last sold, and there is the price that the Inland Revenue, as it then was, would be prepared to accept without a quarrel, and the two are seldom the same. Gold is supposed to be the exception. It is the most uniform of commodities, and one ounce of it is very like another. Yet when it reached its record in the last days of January this year, the spot price touched $5,608 an ounce on one reading, the London benchmark was fixed at about $5,405, and the New York futures contract settled at $5,318.40. Three respectable measurements of the same metal on the same days differed by nearly three hundred dollars. I mention this not to be pedantic, but because it is the first of several reminders in this story that the number on the screen and the metal in the vault are not quite the same thing.

The number on the screen has since fallen a long way. By the middle of October gold was trading at about $4,112, some 27 per cent below its January peak. Silver did much worse. Having touched $121.62 on 29 January, it lost more than half its value and now trades at around $56. For those who have spent three years waiting for the precious metals to stop interrupting the serious business of the equity markets, a comfortable conclusion is available. The whole affair was a trade, the trade has been unwound, and the world can return to its previous arrangements. I think this conclusion is mistaken, and the reason is that while the price was falling, the people who actually hold metal were behaving as if nothing of the kind had happened.

The buyers who did not care

Between April and June, as the gold price fell and the commentary turned sceptical, the central banks of the world bought 288.9 tonnes of gold. According to the World Gold Council, this was the highest second-quarter total on record, and 62 per cent more than in the same period of last year. A quarterly figure does not tell us whether the buying was concentrated in the weakest weeks. But the Council’s reporting shows that these institutions kept adding through a quarter in which gold lost 14 per cent of its value. Speculators chasing a rally do the opposite. They buy as the price rises and sell as it falls. Central banks, which are by temperament the most cautious and price-sensitive investors in the world, bought heavily into a falling market, and they did so without troubling to explain themselves.

Silver spent the year doing something that no price chart could show. The Silver Institute’s World Silver Survey for 2026 estimates a sixth consecutive annual deficit in the silver market, of 46.3 million ounces against 40.3 million the year before, with mine production flat. I should add that some secondary reports of the same survey have given a much larger figure, of 215 million ounces. The two cannot easily be reconciled, and I prefer the Institute’s own headline, which is the smaller. On either figure, the shortfall between what is used and what is produced widened in a year when the price fell by more than half.

That pairing of a violent fall in price with a worsening physical balance is the subject of this article. I am aware that the word “shock” has been worn smooth by journalists, who apply it to anything that moves a chart for a week. I use it here in a narrower sense. I mean the condition in which the price of a metal and its physical availability cease to agree with one another, and the gap becomes too wide for ordinary dealing to close. A spike in price is an event, and events reverse. What gold and silver have been telling us is that the way in which the world settles its accounts is changing, and changes of that kind are considerably harder to reverse than the enthusiasm or panic that accompanies them.

Part of the reason why this has gone unexamined is that precious metals sit awkwardly within the way financial news is organised. They have no product launches and no quarterly earnings. Their most important movements are recorded in technical backwaters such as lease rates, exchange-for-physical spreads and official reserve statistics, and they only become intelligible when several of these are read together. Meanwhile, the attention of the investing public has been directed with great efficiency towards artificial intelligence and a handful of very large American companies, while the foundations of the monetary order have been quietly rearranged by the institutions whose job it is to maintain them.

A note on the evidence

Before going further, I must say something about the quality of the evidence, since a lawyer who does not do this is not doing his job. Some of what follows is documented and can be stated plainly, with its source. Some is inference from the documented facts, and I shall say when I am inferring. A smaller part is speculation, and I shall label that as well.

The central bank figures in particular deserve caution. They are estimates, they are revised, and one of them was revised this year almost out of existence. The World Gold Council’s first estimate of central bank buying in the first quarter of 2026 was 244 tonnes, and it was widely reported as evidence of sovereign conviction on a grand scale. In its second-quarter release, the Council revised that figure to 57 tonnes, citing new data and analysis. That is a reduction of nearly four-fifths, and the correction was published with much less fanfare than the original. In court, a witness who revised his evidence by four-fifths between examination-in-chief and cross-examination would not leave the box with his reputation intact.

Yet this is not quite the case of a discredited witness. Central bank buying is genuinely hard to measure. Reporting is uneven. Some purchases are routed through sovereign wealth funds and state oil funds whose holdings are counted inconsistently, and the line between reserve gold and gold owned commercially by the state is not always clear. What survives the revision is the direction of travel. A pattern documented over four years and from several independent sources is not overturned by the restatement of one quarter. The surveys of the central bankers themselves continue to record a near-unanimous intention to keep buying, and if the accumulation were an illusion of the statistics, the people doing it would have stopped telling us that they mean to continue.

Not 1980, and not 2011

There have been two great precious metals booms in my adult life, and it is worth explaining why the present one differs from both. The boom that ended in 1980 was a genuine monetary crisis. Inflation was in double figures after the second oil shock, and the dollar was in retreat. It coincided with an attempt by the brothers Nelson Bunker Hunt and William Herbert Hunt to corner the silver market with borrowed money. Gold peaked at about $850 an ounce on 21 January 1980. Silver touched roughly $50 and then collapsed in late March, when the Hunts were unable to meet a margin call, on the day still remembered as Silver Thursday. Paul Volcker’s Federal Reserve had by then pushed short-term interest rates towards 20 per cent, and would do so again in 1981. The collapse of silver owed at least as much to leverage as to monetary policy.

The boom that ended in 2011 was what the market calls a policy trade. It rested on quantitative easing and on the fears generated by the European sovereign debt crisis, and it unwound when those policies were reversed or forgotten. Both episodes were financial events with financial causes. What distinguishes the present position is that the price is supported by the balance sheets of institutions with no intention of selling, and pressed upwards by industrial consumption that cannot easily be switched off. That is a harder thing to deflate than a speculative position.

In December 1912, the elder J. P. Morgan was examined by the Pujo Committee of the House of Representatives, which was investigating what it called the Money Trust. Asked about the nature of money, he answered: “Money is gold, and nothing else.” The sentence has been quoted so often that it has lost its force. For most of the past half century it has been treated as the remark of a man from a vanished world. The behaviour of the world’s central banks suggests that the world was not quite so vanished as we were told.

What fell, and what did not

To understand why the spring collapse changed so little, it helps to know what happened in the London market during the winter. The pressure began to build in the last weeks of 2025 and broke in January. At one point silver rose 25 per cent in a fortnight. The gold-to-silver ratio, which is the number of ounces of silver needed to buy one ounce of gold, fell below 45, a level rarely seen outside periods of real stress. Silver did not reach its high of $121.62 by any gradual repricing. It got there through a squeeze, and the squeeze left its usual marks. The cost of borrowing physical silver rose sharply as those who had sold short scrambled to cover. Metal for immediate delivery became dearer than metal for later delivery. The gap between the London price and the New York futures price widened until it was plain to anyone who looked that the two markets had temporarily ceased to price the same commodity.

When a squeeze of that kind unwinds, the price usually collapses, and this one obliged. Silver lost more than half its value between late January and the autumn. Gold, which is more liquid and more widely held by institutions, lost rather less, and the ratio drifted back to about 70, close to its long-term average.

A high price normally ends a squeeze by choking off marginal demand and by drawing out more metal, whether from private hoards or from new short sellers. This year demand was hardly choked off at all, and the hoards stayed where they were. Industrial users kept buying, because a manufacturer of solar panels or electrical connectors does not shut down a production line because one of his inputs costs 30 per cent more than it did last quarter. He absorbs the cost and passes on what he can. Private holders in Asia largely held on to what they had. The work of breaking the squeeze was done almost entirely by the short sellers. Financial traders, trading futures and options, pressed the paper price down until the premium on physical metal disappeared. The collapse took place in the paper market. The metal itself became no more abundant, and two months later the projected deficit for the year was revised upwards.

Official understanding of these matters has never been strong, and one of the most senior economists in the world was candid enough to admit it. In July 2011, when Ron Paul asked him in a House committee whether gold was money, Ben Bernanke said that it was not, and that it was an asset. Two years later, appearing before the Senate Banking Committee, he was asked about the fall in the gold price and replied: “Nobody really understands gold prices, and I don’t pretend to really understand them either.” The two statements sit uneasily together. We are told that gold is an asset whose price nobody understands, and whose monetary function has been permanently retired. That is a fair description of the official view for a generation, and it is why the conduct of the reserve managers this year deserves more weight than any speech. They are acting on a view of gold that contradicts the doctrine their predecessors wrote down.

The bid that does not trade

Central banks have always held gold. Their reluctance to buy it was an aberration confined to the quarter century between the end of the Bretton Woods system and the financial crisis of 2008. What is new is who the buyers are, and their evident indifference to price.

The World Gold Council’s ninth annual survey of central bank reserves, published this year, found that 89 per cent of the central bankers who responded expected global gold reserves to rise over the following twelve months, and that a record 45 per cent expected their own holdings to grow. A survey of reserve managers by HSBC put the share planning to add gold at nearer 40 per cent, and a majority of these said that current prices would not deter them. These are people who are paid to buy cheaply and sell dearly. This year they abandoned the habit, and they did so knowingly.

Research by Morgan Stanley, circulated during the year, found that gold now forms a larger share of the reserves of the world’s central banks than American Treasury securities, for the first time since 1996. The instrument that anchored the post-war monetary order, and that every central bank in the developing world was once told to accumulate as a matter of prudence, has been demoted within the official sector’s own portfolio. It has not been abandoned. The distinction matters, because demotion points to diversification rather than panic.

The buying has been led by institutions with good reason to want some distance from the dollar clearing system. When the Western powers froze Russia’s foreign reserves in 2022, they showed every central bank with any exposure to sanctions what might be done to it. China’s declared holdings have risen steadily. The more interesting Chinese figure is the one that is not published. It is widely inferred, from analysis of trade flows rather than from any official statement, that China’s true holdings are substantially larger than those it reports. I cannot say whether the inference is correct, and nor can anyone outside Beijing. Poland, Turkey, India and Singapore have all added to their reserves, along with a changing group of Gulf and Central Asian states. There is nothing ideological about the pattern. It is defensive, and it is being carried out by governments that have watched the largest participant in the reserve system rewrite its rules to suit itself. As a lawyer, I can only observe that property held under a title its custodian may revoke at pleasure is not property in any useful sense, and that the reserve managers have reached the same conclusion.

One consequence of all this receives little attention. Reserve gold does not trade. It is moved from one vault to another and then left there, sometimes for decades. Every tonne bought by a central bank is therefore withdrawn from the stock available to settle transactions in the market that sets the price. Estimates of the free float in London, which is the metal genuinely available for settlement, lie between 3,000 and 4,000 tonnes. The figure is an inference rather than a published statistic, since the London Bullion Market Association does not report in a way that allows it to be calculated precisely. Against a float of that size, 289 tonnes in ninety days is no rounding error. It is something like 8 per cent of the tradable stock absorbed in a single quarter, by buyers who do not intend to sell it back. This arithmetic should be treated as an indication of scale rather than as a measurement, since not every tonne a central bank buys comes out of the London float. Some purchases are settled with metal that was already lying in a vault and would never have traded. Even so, the floor under the gold price is no longer a level on a chart. It is the standing policy of several dozen sovereign institutions, and as their buying absorbs more of the float, any further rise becomes less a matter of speculative enthusiasm and more a matter of how many tonnes remain to be bought.

In 1966, long before he became Chairman of the Federal Reserve and the most celebrated central banker of his age, Alan Greenspan wrote an essay called “Gold and Economic Freedom”. It contains a sentence that he spent the rest of his career failing to act on: “In the absence of the gold standard, there is no way to protect savings from confiscation through inflation.” Read today, it sounds less like libertarian theory than like a description of what the reserve managers are doing.

The metal that does not come back

Silver is the more instructive of the two metals, because it attracts less commentary. Gold is a monetary object, and monetary objects invite grand narratives. Silver is consumed, which makes it duller to write about and much more revealing.

The difference between them is mechanical. Nearly all the gold ever mined is still with us, in bars, coins, jewellery, official vaults and exchange-traded funds. Very little is ever destroyed. A deficit in the gold market is therefore a transfer between holders rather than a shortage in any absolute sense, and the metal can always be found at a price. Silver does not work in this way. Somewhere between half and three-fifths of each year’s consumption goes into industrial uses in which the metal is dispersed into finished goods in quantities too small to recover economically. A solar panel contains perhaps fifteen to twenty grammes. An electrical connector contains a fraction of a gramme. Pacemakers, car sensors, mobile telephones and the circuit boards of computer servers each carry a trace, and when the product is scrapped the trace usually goes with it to the landfill.

The result is a one-way flow that distinguishes silver from every other precious metal. Roughly half of each year’s mine output leaves the monetary stock for good. What remains continues to circulate, but against a base that is being steadily eaten away. Year after year of deficits does more than tighten the market. It converts a monetary metal into an industrial one, tonne by tonne, until the remaining monetary stock is small enough to be squeezed. That is what happened in January.

It follows that the supply of silver responds hardly at all to price over any period that matters. In a market with spare capacity, a high price brings out more of the commodity. In silver, it brings out a decision to build a mine, and the building takes years. In the meantime, high prices encourage the industries that use the metal to economise on it. This reduces demand slightly, which relieves the deficit slightly, which encourages complacency, which delays the investment that would actually solve the problem. The cycle defeats itself, and it has now been running for six years. That the deficits are real is a matter of record. That they must eventually produce a physical shortage and much higher prices is an inference. It is a reasonable inference, but it depends on the assumption that demand will keep growing and supply will keep failing to respond.

Keith Neumeyer, the chief executive of First Majestic Silver, has spent much of the past decade arguing in interviews that silver ought to trade between $100 and $200 an ounce. He has an obvious interest in saying so, and for most of that decade he was regarded as an eccentric. Having seen the price pass $121 before retreating, he looks less eccentric now than he did in 2016.

A deficit that widened while demand fell

The strangest part of the silver story is also the least reported. The largest single industrial use of silver is in solar panels. Their manufacturers consumed 186.6 million ounces in 2025, and their share of industrial demand rose from about 11 per cent in 2014 to nearly 29 per cent a decade later. Then something happened that the optimists had not fully foreseen. The industry began to economise. Squeezed by collapsing panel prices and anxious to cut costs, manufacturers reduced the amount of silver in each cell, using thinner printed lines, better pastes and alternative methods of making the electrical contacts. Demand for silver from solar manufacturers fell by 6 per cent in 2025, and is forecast to fall by a further 19 per cent this year, to about 151 million ounces. That would be the largest annual fall on record for silver’s largest industrial customer.

On a simple reading, a fall of nearly a fifth in the largest source of demand ought to have cured the deficit. It did the opposite. The projected shortfall for 2026 is wider than that for 2025. The explanation lies in what has grown while solar demand has retreated. Total industrial consumption fell by 3 per cent in 2025, to 657.4 million ounces, and is forecast to fall by another 3 per cent this year, to about 639.6 million. Beneath these totals the composition of demand is changing. Demand linked to the building of data centres, to high-speed electrical equipment, to electric vehicles and their charging networks, and to the modernisation of electricity grids is expanding, and it is expanding in uses where silver is hard to replace. A solar manufacturer can redesign a cell to use less silver. A high-voltage connector or a switchgear contact in a power station has far less room to manoeuvre, because the conductivity of the metal is the whole point of using it.

The market therefore absorbed the loss of some 35 million ounces of solar demand and still ran a wider deficit, because the demand that replaced it was both larger and harder to economise. This is what separates the present cycle from an ordinary commodity correction. The destruction of demand in the largest segment produced a deficit rather than a surplus, in the same year that mine supply was flat and central bank purchases of gold reached a record.

Eleven years between a price and a mine

The usual objection to all this is that high prices will eventually call forth more metal. That is true, and largely irrelevant over any period that concerns a policymaker or an investor, because the delay between a price signal and a tonne of refined silver is measured in years, and sometimes in decades.

Economic deposits have become harder to find. Average ore grades in the main producing regions have been declining for a generation, which means that more rock must be moved and processed to yield the same quantity of metal. Every reduction in grade therefore raises the energy and capital needed for each ounce, at a time when both have become dearer. Environmental permitting has lengthened in most jurisdictions, and in several important ones the interval between discovery and first production now regularly exceeds a decade. Financing costs remain well above the levels of the 2010s, which raises the return any project must promise and kills the marginal ones. On top of all this, the shareholders of mining companies, having watched their money destroyed for twenty years, now insist on what they call capital discipline. Even producers with cash in hand are reluctant to build.

The demand side, by contrast, needs almost no new investment at all. A data centre can be financed and built within eighteen months, and a factory for electric vehicles retooled in two years. Industry can therefore expand its consumption of silver on a cycle perhaps a fifth as long as the cycle needed to expand supply, and it has been doing so since 2021.

The same mismatch explains why price has stopped working as a signal. Shortages are supposed to cure themselves through price, and something of the kind has been happening in silver. It has not been enough, because demand in the principal uses responds weakly to price and supply hardly responds at all. Almost the whole burden of adjustment therefore falls on price, which must move a very long way to achieve a modest rebalancing. Take a solar manufacturer whose silver costs rise by 40 per cent. Silver may be 10 to 15 per cent of the cost of materials in a panel, so the direct effect is a few percentage points. He will redesign his cells to use less, as the industry has done, cutting silver content by a fifth in a single year. But there is a floor. At some point the printed lines become too thin to print reliably, and the chemistry of the paste runs out of room. Having reached that floor, the manufacturer will pay whatever the metal costs, because the alternative is to stop production. On the supply side, matters are worse. The decision to build a mine rests on a forecast of price that must hold for a decade. A price that spikes for eighteen months and then halves, which is what has just happened, is the least helpful signal imaginable. It encourages investment at the top and then punishes it. Mining executives have learned, at great cost to their shareholders, that a high price today is weak evidence of a high price in 2034.

The float, the vault and the lease rate

Anyone who wants to know whether the physical market is genuinely tight should ignore the spot price and look instead at instruments that the newspapers almost never report. Since these are technical matters, I shall explain them as briefly as I can.

The lease rate is the cost of borrowing metal. In a comfortable market it stays close to ordinary interest rates and attracts no attention. When it rises sharply, it means that those who have sold metal they do not own cannot find physical metal to deliver, and are paying heavily to borrow it. In extreme cases this produces the kind of squeeze that drove silver above $121 in January. The forward curve tells the same story in a different way. When metal for later delivery costs more than metal for immediate delivery, the market is comfortable. When immediate metal commands a premium, which the trade calls backwardation, it is the mark of scarcity. Silver went into backwardation during the winter, and returned to it from time to time during the spring.

The exchange-for-physical spread is the most obscure of the three and the most revealing. It measures the gap between the London price and the New York futures price. When that gap grows wider than the cost of shipping and insuring metal across the Atlantic, one market has run short of deliverable metal and the other has not. This happened in silver early in 2021, when the spread widened so far that it briefly threatened the delivery arrangements of the New York exchange. It happened again last winter. The precedent matters because the 2021 episode was resolved by moving a large quantity of existing metal from London to New York rather than by any new supply. That drew down London’s float and left the system more fragile than before. Each resolution of a squeeze has so far made the next one easier, though this is a tendency rather than a law.

These readings matter more than the spot price because the spot price can be moved by paper. A sufficiently determined financial seller can press down the futures price without owning an ounce of metal, and this year that is largely what happened. London’s float was not rebuilt, and the deficit was not closed. What the fall in price removed was speculative positioning, not scarcity. Nor does the fall look to me like the bursting of a bubble. A bubble bursts when an asset passes back from speculators to its natural owners. In gold and silver, the natural owners are the central banks and the industrial consumers, and they have been the marginal buyers throughout. I can see no great body of holders waiting to sell into the next rally, though that is an observation about what is visible rather than proof that no such body exists. It is at least consistent with the fact that the record prices of January did not produce a wave of profit-taking.

Tariffs pointed at metal

Trade policy has added a further pressure, less discussed than geology but more immediately consequential. In November 2025, silver and the platinum-group metals were added to the United States Geological Survey’s list of critical minerals. This opened the way to an investigation under Section 232 of the Trade Expansion Act of 1962, the statute that allows the President to restrict imports on grounds of national security. The market spent the following weeks pricing the possibility of a tariff on refined silver entering the United States. The consequence of such a tariff was obvious. It would pull metal towards America and away from London, which is exactly the mechanism that drained the London float in 2021.

On 14 January this year, the President issued a proclamation declining, for the time being, to impose tariffs on processed critical minerals and their derivatives, and directing his officials towards negotiations with trading partners. Two further proclamations were published in the Federal Register on 20 January. The schedules are dense, but the strategic position is clear enough. Duties were imposed on some categories of goods, while the precious metals were left outside, for now. Silver had already been excluded from the so-called reciprocal tariffs the previous September.

In effect, the largest economy in the world has built a mechanism capable of redirecting the world’s flow of silver at short notice, and has not yet used it. The metal moves freely, and the threat of restriction does the work. If the tariff were imposed, physical silver would be drawn into American warehouses, London would tighten further, and the transatlantic spread would widen. That is the sequence that preceded the 2021 dislocation, but it would now take place in a market with a smaller float and six years of deficits behind it. China, for its part, has spent several years converting its dominance in the processing of critical materials into leverage over those who consume them, by means of export licences, quotas and state allocation. The restrictions change from quarter to quarter, and the reporting of them is imperfect. That uncertainty is itself a cost, because no manufacturer can plan a factory for the next five years around a policy he cannot read.

There is an uncomfortable precedent. On 15 August 1971, Richard Nixon suspended the convertibility of the dollar into gold, and presented the decision as an act of responsible stewardship. In a narrow sense perhaps it was. What he did not say, because it was not yet clear, was that cutting the last formal link between a major currency and a physical metal would begin half a century of competition between financial assets and real ones, with the state permanently on the side of the former. Governments that can create money at will naturally prefer that their subjects should hold claims to money rather than things. The present use of trade law to direct the flow of metals is the same preference expressed by other means.

The bill arriving

Behind all of this sits an arithmetic that no analysis of commodities can alter, because it is a matter of law and demography rather than geology or sentiment. On the estimate of the Committee for a Responsible Federal Budget, the American federal government ran a deficit of roughly $2 trillion in the fiscal year 2026, and paid about $1.1 trillion in interest. That is a record 3.4 per cent of national output, and the second largest item in the federal budget, ahead of both defence and Medicare. The Congressional Budget Office, which measures net interest on its own conventions, puts the figure nearer $1 trillion, and projects a rise from 3.3 per cent of output to 4.6 per cent by 2036, on the assumption that nothing dramatic changes. Budget modellers at the University of Pennsylvania have suggested that merely stabilising the ratio of debt to output would require permanent fiscal tightening of several percentage points of national income, sustained for decades.

I do not claim that deficits make the price of gold rise. Over short periods the relationship is unstable and the correlation unreliable. A narrower claim does survive examination. A government facing rising interest costs and slowing growth has few politically available options. Higher taxes and deep spending cuts are both politically dangerous, and further borrowing merely postpones the reckoning while enlarging it. What remains in the end is the quiet erosion of the currency, which is the mechanism Greenspan described in 1966, and the more deliberately that erosion is pursued, the more attractive becomes an asset that is no one else’s liability.

Ray Dalio put the short version of this to an audience at Davos in January 2020, when it seemed merely provocative: “Cash is trash.” His fuller argument, developed over several years, was that the debt cycle had matured to the point where the main central banks would be forced to finance their governments’ debts with new money rather than allow interest rates to rise, and that holders of currency would bear the cost. Six years later, with interest consuming 3.4 per cent of American output and rising, the argument looks less like a provocation and more like a forecast that has begun to come true.

The great economies

It is customary to speak of “the global economy”, but the phrase conceals very different situations. China’s growth has been tied to strategic commodities for thirty years. Its property market is weak and its households are cautious, which narrows the room for policy. The state’s habit of accumulating physical assets in uncertain times has been copied by Chinese households with remarkable fidelity. Chinese retail demand for gold has surged whenever confidence in the domestic financial system has dipped, and the metal functions there less as an investment than as insurance held outside the banking system. This habit has deep historical roots, and nothing has yet replaced it.

America is in a different position. Its deep capital markets and the reserve status of the dollar continue to give it resilience, though resilience is not immunity. Persistent deficits and rising interest costs have made every fiscal decision more expensive, and political polarisation has made such decisions harder to take. Manufacturers trying to bring production home meet higher costs at every stage, and uncertain access to strategic materials. The project of reindustrialisation is being undertaken just when the materials it needs have become most expensive.

Europe’s difficulty is the most self-inflicted. The European Union has committed itself to an energy transition and an industrial strategy that depend on reliable access to critical materials. It has very little mining of its own, and a permitting regime that makes new extraction slow even by international standards. The result is a bloc structurally short of the very commodities its own policies have made most valuable, competing against Asian producers with much readier access to supply. I take no pleasure in pointing out that the British position is no better, and that our own planning system would turn the opening of any new metal mine into the work of many years and many lawyers.

In March 2022, Zoltan Pozsar, then a strategist at Credit Suisse, published a note arguing that the freezing of Russia’s reserves had broken the assumptions underlying the dollar-based financial order. “We are about to witness the birth of Bretton Woods III,” he wrote, and he predicted that the new order would rest on commodities rather than on credit. Whether the label is apt is open to argument. His substantive claim, that reserves would increasingly be held in things rather than in promises, has been borne out by four years of official conduct.

The machines that eat silver

If the deficits are structural, it is because the demand behind them is not discretionary. The industries that use silver do not buy it because it is fashionable. They buy it because the alternatives perform worse, and in several uses there is no alternative at all. Data centres and the equipment that powers them use silver in contacts, power modules and high-speed connections, and the quantity appears to rise with computing load rather than with floor space. Electricity grids are the least glamorous and probably the most important consumer, since every connection, breaker and transformer contains a little silver, multiplied across an enormous installed base, and grid renewal is now under way in Europe, China and America at the same time. Solar panels use less silver per cell each year but are still the largest single consumer, and their growing volume has so far offset the economies. An electric car contains materially more silver than one with a petrol engine, mostly in its power electronics and high-voltage connections. Defence procurement has expanded sharply and is almost wholly indifferent to price, since no ministry renegotiates a contract for guided missiles because a metal has become dearer. Medical devices use trivial quantities, but there the requirement is absolute.

Most of this demand flows from decisions already taken. A grid renewal programme authorised in 2024 consumes silver in 2027, and an electric vehicle designed in 2023 fixes its bill of materials for a decade. The demand is largely committed, and it was committed before anyone thought carefully about whether the supply existed.

Mr Buffett’s objection

It would be unfair to present all this without the strongest objection to it, and the strongest objection was made by Warren Buffett in a talk at Harvard in 1998. Gold, he said, “gets dug out of the ground in Africa, or someplace. Then we melt it down, dig another hole, bury it again and pay people to stand around guarding it. It has no utility.” His point, which he has repeated in various forms ever since, is that an asset producing no income is worth only what the next buyer will pay for it.

I have always found the argument stronger as a description than as a criticism. What has changed since 1998 is the identity of the next buyer. A large share of buyers are now sovereign institutions whose motives are political rather than financial, and who are therefore indifferent to the absence of income. For them, the fact that gold is nobody’s promise is the whole attraction. Mr Buffett’s men standing around guarding a hole are doing precisely what the reserve managers want done. It is the paper in the other vaults, whose value depends on the goodwill of its issuer, that has begun to look like the speculation.

What the forecasters cannot agree on

Forecasts are usually read for their numbers. They ought to be read for their disagreement, and this year has produced an unusually wide one. Goldman Sachs put gold at $4,900 by the end of 2026. J. P. Morgan’s commodity team went much further, projecting an average of $6,000 in the final quarter and $6,300 by the end of 2027, before a round of downward revisions drew the general expectation towards $4,500 by the end of the year. On silver, J. P. Morgan forecast $63 in the fourth quarter, a level the metal had already exceeded before it fell back. BullionVault’s aggregate of several forecasting models put silver above $80 by December.

The disagreement is one of principle rather than of detail. The bears reason from a monetary model in which persistently high interest rates depress assets that pay no income, and on that model gold above $4,000 is expensive. The bulls reason from a model in which the identity of the marginal buyer has changed, so that the price is simply what is left over when official and industrial demand meet a shrinking float. The two models do not share assumptions, and further data will not reconcile them, though it may force one side to revise its assumptions. My own view is that the bears have the stronger case over the next twelve months and the bulls over the next five years, and that these positions are compatible. A tightening of monetary policy that depresses gold next year will not close the silver deficit, and it will not persuade a single central bank to sell. Much of the public argument about precious metals is really an argument about time disguised as an argument about facts.

How it may end, and how I may be wrong

The least dramatic outcome is also, on the evidence, the most probable. Gold and silver grind higher over several years, interrupted by falls like the one we have just seen. Mining investment eventually responds, new supply arrives in the early 2030s, the deficit narrows, and the metals settle at a permanently higher level without any crisis. This requires nothing more than that central banks continue to buy and industry continues to consume, and both are already happening.

The worse possibilities differ only in degree. A tariff decision, a policy error or an unexpected drain on inventories could produce another squeeze like January’s, but in a market with a smaller float and dealers less able to carry stock. In that case the price would overshoot violently, industrial users would be forced to buy at any price to keep their lines running, and the dislocation would pass into manufacturers’ margins and from there into the price of electronics and building. This requires nothing new, only the repetition of something that has already happened once this year. Beyond that lies a general monetary or fiscal crisis, such as a failed auction of government debt or a currency collapse in a large emerging economy, which would force a sudden change in the way reserves are held. Gold would then rise severalfold, official buying would become a scramble, and the existing reserve system would be abandoned formally rather than informally. I think this the least likely outcome. It is also the one in which the phrase “precious metals shock” would earn its keep.

An argument that cannot be falsified is a temperament rather than an analysis, so I should say what would prove me wrong. Central bank buying would have to fall away, not for one quarter but for several years, and the surveys of intention would have to reverse. A sustained fall below 200 tonnes a quarter across many reporting institutions would begin to test the case, though the revision of a single quarter does not. Mine supply would have to surprise upwards in a way that present project pipelines do not suggest, and that would require a wave of permitting reform across several jurisdictions, which is a political matter and therefore slower and less predictable than geology. Substitution would have to accelerate in the uses where it is now difficult. There is real work being done on reducing silver in solar cells and on copper-silver hybrids in some connectors, and if a genuinely cheap substitute were found for high-voltage contacts, much of the structural case for silver would weaken. Finally, the public finances of the major economies would have to improve, which would require either sustained growth well above trend or a political willingness to raise taxes and cut spending at the same time. I see no sign of either. If one of these conditions were met, the case would survive. If most of them were met within the same decade, it would be close to refuted, and anyone who insists that this cannot happen is overstating the case as badly as those who called the spring collapse the end of the story.

I come back to the three prices of gold in January. The difference between them was a small reminder that the number on the screen is not the metal in the vault. The events of this year have made the same point on a larger scale. The paper price of gold and silver has fallen a long way. The metal itself has continued to move steadily into the hands of those who do not intend to part with it, and out of the market altogether into the machinery of the modern world. When the price and the metal disagree for long enough, it is not usually the metal that gives way.

For the avoidance of doubt, none of the above is to be taken as financial advice. Investments can fall as well as rise, and silver has lately shown how far. Before risking your savings, I advise you to consult someone licensed by your government. He will surely have your best interests uppermost in his mind.

Sources: