Five people, five assets, one broken system
The previous article named nine dysfunctions in how real-world assets move. Named that way they stay abstract — the kind of thing you nod at and forget. So here they are as five ordinary days: a copper cargo, a supplier payment, a household's savings, a bond posted as collateral, and a stake in a fund. Five people, each trying to do something entirely reasonable, each stopped at a precise point by the same missing thing. And then a sixth person, whose asset never gets a day at all.
- None of these six people is defeated by a shortage of capital, an absent counterparty, or a lack of willing buyers. Each is defeated by bookkeeping — by the fact that no two institutions in the chain share a record of what the asset is or what has happened to it.
- Kwame's copper is financed once instead of several times, because no second lender can independently confirm it exists, where it is, or that it has been pledged only once.
- Priya's payment crosses four ledgers to travel between two neighbouring economies, and loses value at every boundary.
- Alice can read the news that repriced her holding twelve hours before her own venue will let her act on it — and the products that would suit her savings carry minimum tickets orders of magnitude above her whole portfolio.
- Lena forgoes yield not because she lacks collateral, but because her collateral cannot be delivered and matched before her counterparty's deadline.
- Sofia's fund stake is performing — and is likely to clear below carrying value, because the only buyers are the ones an intermediary can assemble over months.
- And Fatima's savings — thirty years of gold, entirely legal, unambiguously hers — have no story at all, because nothing about them can be verified without surrendering them. That silence is the ninth dysfunction, and the largest.
In How real-world assets move today we set out nine dysfunctions — markets that sleep, settlement that lags reality, state that does not propagate, a long chain of thin roles, no common system of record, double-pledging, forged provenance, leakage, and frozen wealth — and argued that they are not nine problems but one problem wearing nine costumes: there is no shared, live, verifiable record of what an asset is and what has happened to it.
That is a fair diagnosis and a bloodless one. A dysfunction only becomes real when it is somebody's Tuesday.
1 · Kwame and the cargo that can only be financed once
Kwame trades refined copper. A cargo of 5,000 tonnes leaves a smelter, moves by rail to a port, sits in a bonded warehouse, is loaded, and spends five weeks on the water. Against that cargo his bank extends a facility. From the moment it is drawn to the moment the buyer pays, roughly ninety days pass — and for all ninety of them, that money is doing exactly one job.
Ask why and the answer is not credit appetite. His bank would happily lend against the same cargo three times over if it could sell down the exposure and re-originate. The blocker is that nobody else can see what the bank can see. To bring in a second lender, Kwame's bank must send a data pack, answer questions, and ask the counterparty to take on trust a set of facts it has no independent way to verify: that the metal exists, that it is where the paperwork says, that its quality is as certified, and — above all — that it has been pledged to nobody else. Each of those questions is answered by a document, and every document is a photocopy of somebody's opinion.
The market has learned what that costs. In 2014, duplicated warehouse receipts at Qingdao were used to pledge roughly 400,000 tonnes of copper, aluminium and alumina to multiple lenders at once; Glencore's storage arm alone reported more than 8,000 tonnes of aluminium ingots and 112,000 tonnes of alumina it held receipts against but never received. Commodity-backed lending in the region seized up for months afterwards, and lenders have priced warehouse paper more defensively ever since.
- 3The cargo's state changes constantly — loaded, in transit, discharged, inspected, encumbered — and none of those changes reach anyone who is not on the phone.
- 5The smelter, the warehouse, the surveyor, the shipping line, the bank and the buyer each keep their own record. There is no version anyone can point at and call authoritative.
- 6Because no second lender sees the whole picture, the same metal can in principle be pledged more than once — so lenders behave as if it might have been.
- 7The certificate is trusted to stand in for the metal, and the certificate can be forged.
- 8The margin that pays for all that defensiveness comes out of Kwame's financing cost, and appears nowhere as a line item.
The interesting cost here is not the fraud. It is the discount charged to everyone who is honest. Because the paper cannot reliably be distinguished — a genuine receipt from a duplicated one — it tends not to be priced differently, and the honest trader carries part of the cost of the possibility of the dishonest one. That is the classic dynamic of a market where quality cannot be verified at the point of sale, and it is worth separating from credit risk proper. The Asian Development Bank puts unmet global demand for trade finance at roughly $2.5 trillion; the African Development Bank estimates $74–92 billion of it sits on that continent alone. Not all of that gap is unverifiability — but some of what is booked as credit risk is really the cost of not being able to check.
2 · Priya and the payment that goes to New York to get next door
Priya runs treasury for a textile manufacturer in South India. On a Tuesday morning she instructs a $410,000 payment to a dye supplier two time zones away — a routine trade payment between two economies that trade heavily with each other.
The money does not go two time zones away. It goes to New York.
Everything in that chain is legitimate. Each institution is regulated, each is doing something it is paid to do, and none of them is villainous. But the reason there are four of them is not that the payment needs four institutions' worth of work. It is that no two of them share a ledger, so each boundary needs its own settlement, its own screening, its own reconciliation, and its own fee.
Corporate payments of Priya's size are cheaper in percentage terms than the small transfers where cross-border pricing is actually measured — but they run down the same pipe, and the small end shows what the pipe costs when there is nothing to amortise it against. The World Bank's Remittance Prices Worldwide series has the global average cost of sending $200 across a border at around 6.5%, more than double the 3% target the world set itself, with banks the most expensive channel of all. Two decades of competition, digitisation and political attention have moved that number slowly, because the number is not really a price. It is the cost of four institutions agreeing.
- 4Four institutions in a chain that exists to bridge ledgers, not to add value to the payment.
- 2Instruction and settlement are separated by days; the value is in transit and belongs, operationally, to whoever is holding it.
- 5The same payment is screened, recorded and reconciled four times because no one instance of it is authoritative.
- 8Two FX spreads, several fees, and float income — none of which is compensation for risk Priya asked anyone to take.
The costly part of Priya's payment is not the movement — it is the uncertainty window. For three days she cannot state where the money is; for two of them, neither can her supplier. Both must behave as if it might not arrive: she cannot redeploy the cash, and he will not release the goods. So the payment ties up working capital on her side for three days and on his for two — a cost carried at both ends of a trade that was agreed in a phone call, and one that appears on no fee schedule. That is exactly why it survives: nobody is billed for it, so nobody is motivated to remove it.
3 · Alice, who can read the news but not act on it
Alice is a doctor in Nairobi with about $40,000 in savings, some of it in a US-listed fund she bought through a local broker. On a Wednesday at 04:30 her time — just after the Hong Kong open — a policy announcement materially changes the outlook for the sector her fund tracks. Asian-listed companies in that sector reprice within minutes. She reads about it at 06:30 over coffee. Her fund's venue opens at 16:30 local.
So the sector has moved, publicly and visibly, twelve hours before the instrument she actually holds is allowed to respond. Anyone holding the Asian-listed version of the same exposure has already acted; anyone holding a derivative that trades in an open venue has already hedged. Alice can watch, price the implication precisely, and do nothing. When her venue opens, it opens for everyone at once — and the gap has already been priced in by the people who could trade around it overnight. Her order then settles the following day: the US, Canada and Mexico moved to T+1 in May 2024, and the UK, EU and Switzerland have committed to follow on 11 October 2027, which is the industry describing "one business day" as the frontier.
The sleeping market is the visible half of Alice's problem. The other half is what is not on her menu at all.
- Local listed equities and a handful of local funds.
- One or two offshore ETFs, via a broker with an offshore relationship, at a fee layer she cannot see.
- Bank deposits.
- Short-dated government paper in reserve currencies, at retail size.
- Private credit, which has typically run gross yields in the 8–12% range.
- Income-producing commercial property, other than by buying a whole building.
- Institutional funds, where minimum commitments run from tens of thousands to several million dollars.
The exclusion is not regulatory in origin. Nothing in law says a doctor in Nairobi may not own $2,000 of a warehouse or $500 of a Treasury bill. The exclusion is operational: the unit of ownership is a whole unit because the register was built to track whole units, and the cost of onboarding, custodying and administering a small holder can exceed what a small holding earns anyone. So the shelf is narrow — and the money finds somewhere else to go. The Henley & Partners / New World Wealth Africa Wealth Report puts total private investable wealth on the continent at around $2.5 trillion, with just over 128,000 individuals classed as high-net-worth; a large share of that wealth is administered offshore, on platforms built where the shelf is wide.
- 1The venue keeps office hours; the information does not, and neither do the venues where correlated assets trade.
- 2Ownership and payment move a day apart, so capital is idle between decisions.
- 4Local broker → offshore broker → custodian → sub-custodian, each taking a layer.
- 8That layered fee stack is invisible to her at the point of purchase, which is what allows it to persist.
- 9The assets that would suit her savings exist, are performing, and are unreachable at her size.
Minimum ticket sizes are usually explained as investor protection. A good part of what actually sets them is record-keeping cost. A fund's minimum has to cover what it costs to open, verify, service and report to one more line on a register maintained largely by hand. Where that per-holder cost has genuinely collapsed — index funds, exchange-traded products, retail brokerage, and more recently fractional shares — minimums have tended to collapse with it, and regulation has generally followed the operational change rather than led it. The suitability question is real and separate; it is not usually the thing setting the number.
4 · Lena, who has the collateral but cannot deliver it in time
Lena manages collateral for a mid-sized asset manager. On Thursday afternoon her Singapore counterparty issues a margin call to be met by their Friday morning. She has plenty of eligible collateral: a block of government bonds, held with a custodian in New York.
What she does not have is a way to get it there in time. Delivering a security is not a message; it is a settlement. It needs an instruction from her side and a matching instruction from the receiving agent's side, both submitted before a cut-off, and then a settlement cycle that runs to a value date. Her counterparty's agent in Singapore closed hours before the call reached her desk and will not be back at a keyboard until the deadline is effectively upon them. The chain is not slow because anyone is idle; it is slow because agreement between two books requires both books to be attended, and they are attended at different times of day.
So she does what collateral managers do: she posts cash instead. The cash is accepted immediately, and from the moment it leaves it earns nothing for her fund. The bonds, meanwhile, stay exactly where they are — eligible, unencumbered, and useless.
Lena is not short of collateral. She is short of a way to prove she has it, to someone else's satisfaction, before their morning.
- 1The delivery requires two books to be open and attended at the same moment, and they keep different hours.
- 2The delivery leg and the obligation it satisfies do not settle together, so a buffer has to be posted against the gap.
- 8Yield forgone on cash posted in place of securities — a leak that appears on no invoice and in no fee table.
Collateral is the clearest case of an asset whose usefulness is bounded not by what it is but by how fast its state can be proven to a third party. A bond that cannot be shown to be free, eligible and yours within the counterparty's window is, for that purpose, not collateral at all. Stress events tend to expose two distinct constraints at once — how much good collateral exists, and how fast it can be moved to where it is needed — and the second is far less discussed than the first. Lena's problem is entirely the second: the asset exists, the proof cannot travel fast enough, and the system substitutes cash it did not need to substitute.
5 · Sofia, who needs her capital four years before the fund does
Six years ago Sofia's family office committed to a private fund with a ten-year life. The fund is performing. The problem is that a different commitment has come due four years before the fund's, and she needs part of the capital back.
There is a secondary market for exactly this, and it is not small: private-equity secondary volume hit a record of around $240 billion in 2025, up roughly 48% on the year, of which about $125 billion was LP stakes changing hands. But "there is a market" and "there is a price you can act on today" are different statements. Selling her stake means a data room, a general-partner consent process, a buyer universe assembled by an intermediary, and diligence on a portfolio the buyer cannot see directly. It is a months-long process, not a trade.
And when it clears, it is likely to clear below carrying value. Average LP portfolio pricing in 2025 was 87% of net asset value, though the spread by strategy is wide — buyout stakes around 92%, real-estate stakes closer to 70%. Wherever in that range Sofia lands, the shortfall is not a judgement that the underlying assets are worth less. It is the cost of needing her money at a moment the structure did not anticipate, in a market where the buyer has to rediscover the portfolio from scratch.
- 2No continuous settlement, therefore no continuous price — only a negotiated one, months later.
- 4Intermediaries assembled ad hoc for each transaction, each pricing scarcity as much as risk.
- 8A discount to NAV, part of which is the buyer's cost of verifying what they are buying rather than the risk of holding it.
The secondary discount is usually read as an illiquidity premium — compensation for tying capital up. Part of it is. But part is better described as an opacity premium: the buyer is pricing the cost of diligence plus the residual risk that the diligence missed something. Those are different costs with different cures. Illiquidity is cured by more buyers. Opacity is cured by a verifiable record. The distinction matters because the market has been adding buyers for a decade — volume grew 48% in a single year — and average pricing has not converged on NAV. Something other than the buyer count is holding the gap open.
And the sixth person, who has no story
Fatima runs a small tailoring business in Cairo, and her family's savings are where her mother's and grandmother's were: in gold — bangles, coins and two small bars, bought gram by gram over thirty years, kept at home. She has an asset. It is entirely legal, fully paid for, and unambiguously hers. Nothing has ever happened to it.
It earns nothing, year after year, while inflation works on everything around it. When a supplier demanded payment before a big order and Fatima needed cash for three months, the only bid was the neighbourhood dealer's — and his price assumed the worst about her gold, because he had no way to know its fineness without melting it, and no receipt she holds means anything to anyone but the shop that issued it. She was offered a fraction of what the metal was worth, to be bought back at a premium, and she declined. She cannot pledge the gold without physically surrendering it to someone she must simply trust. She cannot sell a part of a bangle. She cannot insure the hoard at a sensible price, because the insurer cannot verify what is there either. So the savings sit — safe from the bank, and useless to her.
The textbook example of frozen wealth is unregistered land — the trillions of dollars of "dead capital" that lenders cannot touch because title is unclear. But that wealth unfreezes only when the law of the land catches up, and Fatima cannot wait for a legislature. Her version of the problem needs no new law at all. Household gold is the largest pool of private wealth that is frozen purely by verification: the World Gold Council estimates Indian households alone hold on the order of 25,000 tonnes — the largest private gold stock in the world, comparable to the official reserves of the largest central banks combined — and estimates of Turkish "under-the-mattress" gold run to several thousand tonnes more. Egypt, Vietnam, Indonesia and much of the Gulf tell the same story. Where the verification problem has been solved — assayed, vaulted, receipted — gold becomes ordinary collateral the same day: India's regulated gold-loan industry lends hundreds of billions of dollars' worth against jewellery precisely because a bank branch can weigh and assay it on the spot. Everything between Fatima and that outcome is bookkeeping, not law.
- 5There is no record of what she holds — weight, fineness, existence — that any third party can rely on, so no third party will transact.
- 7Purity can only be established by a trusted assay, repeated from scratch by every counterparty, every time; hallmarks and shop receipts can be faked, which prices even the genuine ones as if they were.
- 9The asset supports nothing: no loan at fair value, no yield, no fractional sale, no insurance. It is worth what it is worth, and does nothing.
That is the ninth dysfunction in something close to its purest form. The other five lose margin at the edges of transactions that still happen. Fatima loses the entire financial life her asset should have had — and there is no dramatic failure to point at, just an absence, repeated across hundreds of millions of households whose savings are real, legal and unread.
The most expensive consequence of an unreadable asset is not a bad trade. It is the trade, the loan, the insurance and the product that were never possible in the first place — and therefore never counted.
Six times, the same wall
Lined up, the six stories resolve into one table.
| # | Dysfunction | Whose day it ruined | What it actually cost |
|---|---|---|---|
| 1 | Markets that sleep | Alice, Lena | Twelve hours of unusable information; a margin call met in cash because two books were never open at once. |
| 2 | Settlement that lags reality | Priya, Alice, Lena, Sofia | Working capital consumed in transit at both ends of every trade; buffers posted against the gap. |
| 3 | State that doesn't propagate | Kwame | Nobody downstream knows where the cargo is or what has been pledged against it, so nobody downstream will lend. |
| 4 | A long chain of thin roles | Priya, Alice, Sofia | Four institutions to cross one border; four layers between a saver and an asset; an intermediary assembled from scratch for each private-market sale. |
| 5 | No common system of record | Kwame, Priya, Fatima | The same facts verified, screened and reconciled once per participant, forever — or, in Fatima's case, not verifiable at all. |
| 6 | Double-pledging | Kwame | A defensive discount charged to the honest, because honest and duplicated paper look alike. |
| 7 | Forged provenance | Kwame, Fatima | Certificates trusted to stand in for the asset, and nothing forcing them to stay true. |
| 8 | Leakage | All five with a transaction | Spreads, lifting fees, float, forgone yield, a discount to NAV — none of it on any invoice. |
| 9 | Frozen wealth | Alice, Fatima | Assets that cannot be reached at retail size; assets that cannot be reached at all. |
Notice what is not in that table. Not one of these six people is blocked by a shortage of capital, an absence of willing counterparties, or a market that has judged against them. Kwame's bank wants to lend more. Priya's supplier wants the money. Someone would happily sell Alice $2,000 of a warehouse. Lena's bonds are eligible. There were $125 billion of buyers last year for the kind of stake Sofia holds. A lender would take Fatima's gold as security tomorrow, at fair value, if its weight and fineness could be shown without a trip to the melting pot.
In every case the deal is wanted on both sides and dies in the middle — at the point where one party has to prove to another, quickly and cheaply, what the asset is, who owns it, what has been done to it, and what may be done to it next.
It is worth stating plainly what these six stories have in common, because it is easy to mistake it for a technology complaint. None of them needs a faster network. Messages between these institutions already move in milliseconds. What is missing is not speed of communication but a shared object to communicate about — one record of the asset that every participant can read, that updates when the asset's situation changes, and that a regulator, a lender and a buyer can each rely on without re-verifying from scratch. Real reforms have landed on the books themselves: dematerialisation removed the certificate, central counterparties netted the exposures, delivery-versus-payment removed one leg of settlement risk. But each of those changed one institution's book or one market's mechanics. None of them gave two institutions the same object to look at, which is why the reconciliation function they were meant to shrink is still there.
What comes next
The obvious question is what these six days would look like if the missing thing existed. Not incrementally better: differently shaped. Kwame's cargo financed three times on its journey rather than once. Priya's payment settling in one hop against a record both banks already share. Alice holding $2,000 of a warehouse and $500 of a Treasury bill and moving between them at eleven at night. Lena's bond and her counterparty's claim moving in the same instant. Sofia selling a fifth of her stake against a price on a screen. Fatima's gold earning a yield in a vault instead of sleeping in a drawer.
And the more interesting half of the question is not about speed at all. Everything above is a story about things that happen too slowly or too expensively. The next article is mostly about things that do not happen at all — products there is obvious demand for, that nobody can build, because the asset's state cannot be read and enforced by a machine: the loan sized to a cargo's position, the savings product assembled from five markets, the collateral that manages itself. Faster days are the small prize. The positions nobody can take today are the large one.
That is the next article. The one after it comes back to this table and maps each of the nine, one by one, to the specific mechanism that removes it — because "tokenization fixes it" is not an argument, and the nine deserve nine answers.
- How real-world assets move today — the ecosystem problem
- Five people, five assets, one broken system — the problem made real (you are here)
- Imagine the other world — financial Legos and the positions nobody can take today
- Nine problems, nine answers — what tokenization actually fixes
- Rules change. Should the asset? — the new problem tokenization creates
- The Universal Asset State Machine — Decibel Labs' answer
- One machine, five assets — the state machine tested against real asset classes
- Three planes — compliance, physical, economic
- The Universal Asset Token stack — making state enforceable