The Queue Outside The Bank

Families have been saving since the dawn of time. What they have saved in has shaped their fate significantly, whether they realised it or not. Over the years the medium for their savings has varied: cattle; grain in sealed jars. Salt, cloth, and shells where metal was scarce. Silver by weight in Mesopotamia; rings and bangles worn in India. Land, classic really: the olive groves and vineyards that took a generation to mature and were then passed down through each generation that learnt how to cultivate them. Roman families held coin, grain stores, and urban rent. Medieval families endowed their armour, jewels, livestock, and the wealthy saved in annuities (promises of fixed payments for life). Later came government bonds, including the British consol that paid three per cent for a full century, then bank deposits, life insurance, and now, probably the most popular worldwide, the 'house' and the pension fund.

Gold has been the longest lasting of these, and the only one that served as both a global savings medium and a unit of account. Over time it became concentrated in banks, and the paper receipts representing that gold held their value right up until queues started forming outside those banks to redeem it.

This essay is about those queues: where they came from, what they did to the people standing in them, and why a family holding Bitcoin for the long term should understand them well.

How people came to accept receipts as money

The story usually told is that goldsmiths took in gold for safekeeping, issued receipts, and the receipts began to circulate.

In 1640, King Charles I, who was short of funds, seized around £130,000 of bullion that London merchants had deposited for safekeeping in the Tower Mint. He later returned it as a forced loan, but people had begun to realise that the Crown's vault was not safe for their gold. Merchants moved their metal to the goldsmiths of Lombard Street, private tradesmen with strongrooms who owed no duty to the king. By the 1660s the goldsmiths' deposit receipts had become transferable and were changing hands as money in their own right. The goldsmith paid interest on gold deposits while lending the metal out at a higher rate, and kept enough on hand to meet the withdrawals he expected. Fractional reserve banking in England begins here, emerging from the free market.

Amsterdam had done something similar about half a century earlier. The Bank of Amsterdam, founded in 1609, took deposits of coin from citizens and kept a ledger that tracked each depositor's balance. This proved extremely popular because it was so efficient: a balance on a ledger could be transferred with a note rather than a physical delivery of coin, and it traded at a premium over the metal itself.

Then, in the 1780s, that same bank lent its reserves under the radar to the Dutch East India Company. The secret got out, depositors panicked and rushed to take their gold out, and the premium on those ledger entries collapsed. The pattern was set early. The receipt is more convenient than the metal, right up until the day your money is not behind that vault.

The convenience was real. Gold is heavy, conspicuous, and dangerous to keep. A merchant who held his working capital in coin had to guard it, count it, and carry it. The bank took that burden away. Vendors accepted the paper because the bank's name was well known, and because a note could be checked in a moment where coin had to be weighed and verified. Slowly, through each merchant taking the easier option, the paper standard became the norm.

The gold century

By 1717 Britain was on gold in practice, through an accident of the mint ratio set by Isaac Newton. By 1821 it was on gold through law. For the next ninety-odd years the pound was convertible into a fixed weight of metal on demand, and between 1871 and 1900 the rest of the trading world followed: Germany, the Scandinavian countries, the Dutch, the United States by 1879, Austria in 1892, Russia and Japan in 1897, Argentina and India by the turn of the century. For roughly thirty-five years, from 1880 to 1914, almost every economy that traded significantly settled its trade in the same metal.

What the gold standard actually looked like

During the classical gold standard period, roughly 1870 to 1914, the world's stock of monetary gold grew at about 1.5 to 2 per cent per year. Most of that new supply came from the South African and Alaskan discoveries after 1886.

Now look at what that did to the price of money. British government bonds paid around 3 per cent for about a hundred years. A pound in 1914 bought roughly what a pound had bought in 1815. A family could lend at 3 per cent, or borrow at 3 per cent, and be reasonably confident that the unit they would be repaid in, or would have to repay in, would still mean the same thing decades later. That is a very different environment from the one we live in. There was a greater sense of certainty in the medium people traded in and valued in.

The gold century had deflationary decades, agricultural collapses, and a string of banking panics, some of which I will come to below, but they stayed local and revealed bad practice quickly. It was not a stable, gentle period. It did have one thing our current system lacks: an honest price for time and capital. When capital cannot be created out of thin air by a bank, it has to be earned first, saved, and then lent out. The projects that attracted that capital were the ones the market reckoned could realistically pay it back over a long horizon: railways, shipping, telegraph lines, sewer systems. This is long-duration infrastructure, built by people with low time preference. Those willing to risk their capital by lending faced a much larger opportunity cost than today's lender, who today holds a quickly depreciating money and therefore has a much stronger incentive to lend it out on riskier and less sensible ventures. Because the opportunity cost of lending was larger under the gold standard, the incentives led towards a more rational allocation of capital, and therefore, in general, more productive capital.

Where the gold lay

A lot of people picture the gold standard as gold physically moving around the world to settle trade. This is mostly not what happened. Trade settled through bills of exchange drawn up in London. Exchange rates moved inside a narrow band, the so-called gold points, which were set by the market and roughly equalled the cost of shipping and insuring the metal. Physical gold only changed location when the rate moved far enough to justify paying for that transport and insurance, or when fresh production came in from the mines. Over time, central banks simply kept their gold in each other's vaults and transferred ownership by relabelling bars. The New York Fed was still doing this well into the second half of the twentieth century.

This part is crucial: the Bank of England's own reserve was thin. Walter Bagehot, writing in 1873, described the entire British credit structure as resting on a small film of gold lying in Threadneedle Street. The whole thing kept working because, most of the time, nobody was asking for their metal back. Convertibility was a promise, and the promise held precisely because it was rarely tested.

That is the key idea. A claim on gold is not gold, even when you can see the ounces beside your name. It is only a promise that someone will hand you gold on request. And the promise only works as long as not too many people make the request at once.

What happens when the promise gets tested

We will go through five episodes briefly, to show the pattern of these promises being tested.

In February 1797, a small French force landed at Fishguard in Wales. It surrendered within two days, but news of the landing reached the country towns before news of the surrender did. Depositors ran on the country banks, the country banks went to redeem their gold from the Bank of England, and within a week the government suspended gold payments. That suspension lasted twenty-four years.

In December 1825, a boom in Latin American mining shares turned into a collapse. Around sixty country banks failed, and the Bank of England came within a day of running out of coin. By the Bank's own account, it was saved when someone found a box of unissued one-pound notes in a store room and put them into circulation. The public was willing to accept paper from an institution it still trusted, while losing trust in all the smaller ones.

In October 1907, the Knickerbocker Trust Company in New York was rumoured to be exposed to a failed attempt to corner the copper market. The rumour produced a panic, and queues formed outside the trust companies across the city. There was no central bank at the time. J.P. Morgan locked the leading bankers in his library until they agreed to pool their reserves. The Federal Reserve was created six years later, ultimately centralising capital under the banner of protecting other institutions.

In August 1914, war was declared, and the first thing people did, in London, Paris, Berlin, and New York, was ask for their gold. Every major country suspended convertibility within weeks. The classical gold standard, which had taken thirty-five years to build, ended in a fortnight.

In the summer of 1931, Creditanstalt, the largest bank in Austria, failed abruptly. The panic moved to Germany in July, then spread to London, where foreign depositors withdrew gold faster than the Bank of England could sustain. On 21 September, Britain left the gold standard it had rejoined only six years earlier. Eighteen months after that, the United States closed its banks for a week and made private gold ownership illegal, punishable by up to ten years in prison.

So what is common to all five? A run starts when a depositor believes that other depositors are about to run. The bank's actual solvency matters less than what the queue outside implies, because a fractional reserve bank is, by design, unable to pay everyone at once, and the people who understand this best are the ones who get to the front of the line first. Bad news is the spark. War is the most reliable spark of all, because war is exactly when the state most needs the metal and the citizen least trusts the state to leave it alone.

Whenever the queue formed, someone at the back of the line, or even the middle, was left empty-handed.

Why hard money keeps coming back

It is worth remembering that paper money is not new. Yuan China ran a pure fiat note for the better part of a century before over-issuance destroyed it. The American colonies, revolutionary France, and eighteenth-century Sweden each tried an inconvertible currency, and each inflated it away. The lesson Europeans drew from that record was simple, and I think it was correct: a money that can be printed at will, will, sooner or later, be printed.

The return to hard money, when it happened, was always slow and usually for similar reasons. A state that had suspended payment on its promises could not borrow hard money cheaply again until it resumed, because its lenders had just watched it break a promise. Britain in 1821, the United States in 1879, Germany and France in the 1920s. In every case, the government went back to a hard money standard because the cost of not being credible had grown larger than the cost of tightening its belt and refusing to inflate the currency away. The pegs broke under war and panic. They were rebuilt under pressure from those no longer willing to lend. Nobody went back to gold out of sentiment. They went back because their creditors demanded it.

How Bitcoin fits into this

Bitcoin shares with gold the one property that made gold worth queuing for in the first place: it is very difficult to produce more of it. There are twenty-one million coins, and nobody has the ability to print more just because the economy needs them. In that specific sense, gold and Bitcoin are the same type of asset, and the same history applies to both.

What has changed is the arrangement built around it, and it has changed in a direction a seventeenth-century goldsmith would recognise immediately. As Bitcoin has moved from a speculator's asset or hobby to something closer to a reserve asset, a growing share of it is held the way merchants held their gold in Lombard Street: with a third party. Exchange-traded funds and similar vehicles now hold roughly 7 per cent of all Bitcoin, and most of that sits with one custodian. Centralised exchanges hold somewhere between 11 and 13 per cent. Governments hold around 1.5 per cent, most of it acquired through seizure rather than purchase, although the sovereign reserve conversation has clearly started. Add it up and something like a fifth of the supply is an IOU: a line in someone else's ledger, redeemable on request, most of the time.

The reasons families choose the IOU are the same reasons the merchants did, and I want to be fair to them. It is easier, or at the very least, more familiar. It fits inside an existing brokerage account. There is little to learn, and if it is a trusted institution it feels safe. For many holders those are decisive factors.

But the receipt also carries the same risks the merchants' receipts carried, in modern form. The custodian can be hacked (as the Liquid network was in September 2026). It can fail. It can be frozen by a court or a regulator. It can lend the underlying asset to someone who does not return it. It can issue more claims than it holds. Every one of those things has already happened to at least one Bitcoin custodian, and several have happened more than once.

The failures do not arrive at random. They cluster in the downturn, not in the rise. The reason is mechanical. It is during the rise that a custodian is tempted to lend out what it holds and issue claims against what it expects to have. It is in the fall that depositors ask for it back. Mt. Gox in 2014. The lending platforms and FTX in 2022. Each came after an expansion, at the moment the queue formed. This is the 1825 pattern, the 1907 pattern, the 1931 pattern. Same mechanism, different asset.

Where the analogy stops holding

Here the comparison with gold breaks down.

Gold is slow. A run on a gold bank took days to spread and weeks to resolve, and the sheer friction of moving metal gave the bank, and often the government, time to act. Bitcoin settles in about an hour and moves at the speed of an electronic message. A run on a Bitcoin custodian can empty it over a weekend, if withdrawals remain open. The largest failure so far went from rumour to insolvency in roughly three days. There is no lender of last resort, and few willing insurers for the largest custodians. So the discipline that gold imposed on the goldsmith is imposed on the Bitcoin custodian faster, and, as far as we can tell, no custodial arrangement in this asset has yet gone a full decade without a failure.

The digital nature, though, is what makes holding it directly a realistic option for a family in a way that was never quite true of gold. Bitcoin held directly has no vault to guard, no bar to verify, and no weight. It can be structured so that no single key, no single person, and no single location can lose it: two keys of three, or three of five, spread across people the family trusts and places the family controls. It can cross a border as a memorised phrase. And with the right documentation, it can be passed to a spouse or a child who was never technical, because the instructions can be written in plain language and rehearsed while everyone is still alive and well.

None of that is free, and I do not want to pretend otherwise. Direct custody takes on responsibility that the IOU of an ETF does not. Keys get lost. Backups get misplaced. Families fall out. Heirs get defrauded. The trade-off is real and should be stated: the custodian removes the operational burden and adds counterparty risk; holding directly removes the counterparty and adds the burden. What I would argue is that history shows counterparty risk in an asset like this cannot truly be assessed, and that it tends to be low in probability but severe when it lands. It is what we have learnt from history. There is no lender of last resort.

There is a broader shift going on that I think supports this reading. Since 2022, when a major central bank's foreign reserves were frozen overnight, central banks around the world have been buying gold at roughly twice their previous pace, and increasingly they have been repatriating it, moving it out of the vaults in London and New York and into their own. Sovereigns are relearning that a reserve held with someone else is a reserve held at that someone's discretion. In a more divided world, assets that can be held directly are valued more, and ledgers you do not control are trusted less. Families are not so different from states on this point. If anything, they are a little ahead, and they have all the tools they need. As our main way of storing our savings transitions from gold to Bitcoin, we will see self-custody continue to grow in popularity.

For a family that has decided to hold its own keys

If your family has looked at where Bitcoin sits in the world and decided to hold it as long-term capital, the historical record points to a fairly clear conclusion. Queues will form outside these custodians. They have formed under every custodial arrangement for hard money since the goldsmiths first took in the bars. They form faster in this asset than in any that came before it, because of its digital nature and how interconnected we all are. And the only depositors who were never in the queue were the ones who held the asset itself.

Holding it yourself, and holding it well, is not a single or quick decision. It needs structure: how many keys, held by whom, kept where, backed up how, documented how, and rehearsed how often. There are perfectly good arrangements where the family holds most of the keys and a specialist firm holds one, so that no single party, including the firm, can move anything on its own. Those arrangements suit some families and not others.

Schelling Point's work sits alongside that. We never hold client keys and never take custody, not even briefly. What we do is accompany a family through the whole transition: designing the structure, setting it up so it is secure, usable, and survivable, writing the documentation a spouse or an executor could actually follow, and staying with the family over the years as circumstances change. The goal is that the wealth serves the family's life rather than sitting there as a low-level source of worry, and that it passes on cleanly when the time comes.

If you would like to talk through your family's situation in confidence, we would be glad to arrange a conversation.

Schelling Point is a boutique advisory firm helping individuals and families hold Bitcoin as long-term, multi-generational capital. We design custody, inheritance, and continuity structures that are secure, usable, and survivable, so meaningful wealth can be held directly, operated confidently, and passed on cleanly. Read our guide to what Bitcoin estate planning is or how multi-vendor multisig works.