Warehouse receipts were meant to bank Africa's farmers. Mostly, they bank its traders.

Zambia has, on paper, one of the more complete warehouse receipt frameworks in the region. A commodity exchange, ZAMACE, has existed since 2007. The Agricultural Credits Act has provided the legal basis for receipts since 2010. An electronic silo certificate system, running on the same software that supports South Africa's grain market, went live in 2015. By 2018, at least eight private warehouse operators with more than 300,000 tonnes of storage had been certified. In 2020 the receipts became tradeable on a platform launched with the Lusaka Securities Exchange.

Every institution the textbook asks for exists. And yet the Indaba Agricultural Policy Research Institute found that in 2018 only about 4,000 tonnes of maize traded on the exchange, and that use of the electronic receipts was negligible.⁴ Certified capacity of more than 300,000 tonnes; measurable use, close to nil. That figure is from 2018, the most recent independent utilisation data we could find, and we located nothing more recent in the public record, so the current picture is dated even if nothing since points to a step change.

This is worth understanding, because the warehouse receipt is one of development finance's most persistent ideas. The pitch is elegant. A smallholder farmer has no land title, no salary slip, no collateral a bank will accept. But she has grain. Deposit that grain in a certified warehouse, receive a receipt stating its quantity and quality, and the receipt becomes something a bank can lend against. The crop itself becomes the collateral. Less than three percent of formal credit in emerging markets reaches agriculture, and where storage is poor, 30 to 40 percent of a harvest can be lost.³ One instrument appears to address both problems at once. The World Bank, the African Development Bank, FAO and a long list of bilateral donors have backed versions of it across the continent.

So who actually uses these systems, where they exist?

The best answer comes from Malawi, where researchers at IFPRI obtained the complete transaction records of the Agricultural Commodity Exchange, ACE, from 2011 to 2018. Across eight years, the system issued 710 receipts. Large traders and processors deposited the most grain by far. Farmers, most of whom were medium to large scale by Malawian standards, accounted for roughly five percent of the volume. Farmer associations added about five percent more.¹ The instrument designed around the smallholder was used, overwhelmingly, by everyone except her.

Malawi is not an outlier. In Ghana, most receipts issued under the Grains Council system in its early years were issued by certified warehouse operators to themselves, so that they could borrow against their own stock. Ethiopia's exchange, once the continent's flagship, saw its receipt trade dwindle after its state warehousing enterprise was folded back into it in 2017. The one African system that moves serious volume is South Africa's, where electronic silo certificates back around three million tonnes of grain a year. It works because its depositors are commercial farms and trading houses: large, banked, and already legible to the financial system.

The pattern is too consistent to be an accident of one country's design. Something structural is going on, and the Malawi data shows what.

First, a receipt is not a piece of paper. It is a chain of trust. For the paper to mean anything, the grain must be graded to a standard everyone accepts, the warehouse must be certified and inspected, the stock must be insured, a collateral manager must vouch that it is really there, a bank must be willing to treat the paper as security, and a court must be expected to enforce it if anything fails. Every link in that chain is a service, and every service has a price: grading fees, storage fees, insurance, inspection, interest. In Malawi, collateral management alone cost thousands of dollars per warehouse per month, a cost that lands on whoever deposits.

Second, those costs behave like a fixed toll, and tolls punish small loads. Among repeat users of the Malawi system, the average maize deposit was around 24 tonnes. A typical smallholder's marketed surplus is a fraction of that, often well under two tonnes. Spread the toll over 24 tonnes and it may be bearable. Spread it over ten bags and it consumes whatever the farmer hoped to gain. Mario Miranda and colleagues, writing in Agricultural Economics in 2019, reached exactly this conclusion: the transaction costs and risk transfers built into warehouse receipt financing undermine its value to smallholders, which is why, where it exists, it is used almost exclusively by traders, processors and exporters.²

Third, the credit that the receipt is supposed to unlock can be dear enough to defeat the purpose. In Malawi over the study period, borrowers against receipts paid interest of roughly 25 to 42 percent a year. Counting storage fees, insurance and interest, about half of all deposits lost the depositor money, and deposits taken with financing performed worse on average than deposits taken without. The IFPRI authors are blunt: financing costs can make or break a warehouse receipt system. An instrument invented to connect farmers to banks fails quietly when the price of the connection exceeds its worth.

Fourth, and least discussed: the receipt assumes a depositor the system can already see. To deposit, a farmer must be known: her identity established, her grain assessed, her volumes recorded. The warehouse receipt does not create that legibility. It consumes it. South Africa's system works because its depositors were legible before the first certificate was printed. A smallholder in Petauke is asked to enter a formal system carrying documentation the informal economy never gave her. Regional evaluations add a further deterrent that affects every depositor equally: unpredictable trade policy, including the export bans that several countries in East and Southern Africa have applied over the years, makes holding grain in a formal system a gamble on rules that can change mid-season.

Put these four together and the conclusion is not that warehouse receipts are a bad idea. It is that they are a top floor built before the ground floor. The instrument works where depositors are large, banked, informed and already trusted. It stalls where they are small, scattered, unbanked and unrecorded, which is precisely the population it was marketed as serving.

The interesting question is what the ground floor looks like. The cases where receipts genuinely reached smallholders share one feature: an intermediary carried the trust. In Tanzania's coffee sector, cooperatives aggregate members' crop and hold the receipt collectively; studies in Mbinga district found credit accessed this way raised farm productivity.⁶ In Ghana, a donor-supported programme that stitched rural banks and savings companies into the system brought more than two thousand smallholders and small enterprises into receipt-backed borrowing with zero recorded defaults, though the sums involved, a few hundred dollars per borrower on average, show how early that work still is. The IFPRI team's own modelling for Malawi found that farmer groups did best when they aggregated deposits and avoided expensive financing. In every version that works, someone the farmer already trusts stands between the village and the paperwork, bundling small volumes into bankable ones and small identities into a track record.

That is worth sitting with, because it inverts the usual order of reform. The standard prescription runs: pass the law, certify the warehouses, launch the platform, and adoption will follow. Two decades of evidence suggest the opposite sequence. First comes the local institution a farmer can walk to and trust, one that records who she is, what she delivered and when she was paid. Then aggregation, which solves the arithmetic of small volumes. Only then does a receipt system have something to certify.

A warehouse receipt does not create trust. It presumes trust that has to be built somewhere else first.

Zambia is about to test this question at scale. The country has just recorded its largest maize harvest, a projected 4.9 million tonnes by the official crop forecasting survey, and a surplus of that size makes storage, and the institutions around storage, matter more than at any point in a generation.⁵ The certified warehouses exist. The law exists. The exchange exists. On the evidence assembled here, the harder gaps sit underneath and alongside them: the deposits too small to carry the fixed costs, the finance too dear to leave the depositor ahead, the trade policy too unpredictable to hold grain against, and the plainest of the four, the village-level record that would turn an invisible farmer into a depositor a system can recognise at all.

That last gap is where our own work sits. DNC buys maize directly from smallholder farmers in Eastern Province, is building village-scale storage there, and records every purchase against a farmer's verified identity. We did not set out to compete with the warehouse receipt system, and we are far too small to be its answer. Our own reading, drawn from the field as much as from the data, is that the layer most often missing is not certified sheds or enabling laws, both of which Zambia already has, but the trusted local record that makes a smallholder legible to a formal system in the first place. We hold that as an interpretation, not a proven result: small deposits, dear credit and unstable trade policy are real constraints too, and no single study names the record as the dominant one. But it is the layer we can watch being built, one purchase at a time, and where it exists, the paper above it has a better chance of meaning what it was always supposed to mean.