Tokenized Treasuries: Asia's Real Infrastructure Test Sits at the Settlement Layer

Title: Tokenized Treasuries: Asia's Real Infrastructure Test Sits at the Settlement Layer

March 20, 2026, New York. BlackRock announced that its first tokenized fund, BUIDL, had gone live on the Ethereum network, with Securitize providing the technology. Within four months the fund crossed 500 million dollars in assets. That week, most Asian business pages carried nothing. There was no dramatic headline in it. For anyone who works on bank back-ends, though, the real story was never the size of the fund. It sat one layer below. A tokenized treasury bill is a proposal for a new settlement rail. Asia's banking plumbing is weakest at exactly that layer.

Tokenized treasury funds are not new. Franklin Templeton launched its BENJI fund on the Stellar network in 2026, mainly to keep mutual fund share records on-chain. In 2026 the arithmetic changed. On January 10, the US Securities and Exchange Commission approved eleven spot Bitcoin exchange-traded funds. On April 20, Bitcoin halved at block 840,000, cutting the block subsidy from 6.25 to 3.125 BTC. Institutional capital began treating blockchain as an investable product rather than a pilot.

Asia's picture is messier. Hong Kong's Securities and Futures Commission approved spot Bitcoin and Ether exchange-traded funds on April 15, 2026, with trading starting April 30. The Reserve Bank of India began its wholesale e-Rupee pilot on November 1, 2026, and the retail pilot on December 1, 2026. Japan amended its Payment Services Act in June 2026 to define stablecoins legally. Singapore's Monetary Authority launched Project Guardian in 2026 to test tokenized funds alongside decentralised finance. Project mBridge, led by the Bank for International Settlements, started in 2026 and reached a minimum viable product in 2026 with China, Hong Kong, Thailand and the United Arab Emirates participating.

These events look separate. They point at one question: if ownership of a digital asset sits on-chain, in which currency, under which rules, and on whose balance sheet does its value settle?

Answering that requires separating two layers. The first is the record of the asset. Blockchain works here, and works well. The second is the cash leg and final settlement. This is where the arithmetic fails.

The binding constraint on tokenisation projects is not technology speed. It is the legal character of the cash leg. Most cash circulating on-chain today is stablecoin, meaning commercial bank liabilities rather than central bank money. Delivery-versus-payment can in principle happen in a single transaction, but one leg rests on a liability that sits outside the central bank. The repo market, interbank call money and monetary policy transmission all stand on central bank reserve money, and that layer is still not on-chain.

Asia's pilots have made this gap visible. The e-Rupee pilot tried to put central bank money on-chain, but participant numbers and usage depth remain limited. Hong Kong's ETF approval opened the door to institutional demand, yet settlement still runs through conventional clearing. Project mBridge showed that cross-border multi-CBDC settlement is technically possible, but every participating jurisdiction had to change its own legal framework, and momentum stalled there.

Over the past few years I have built a habit of reading on-chain ledgers and pilot reports side by side, and attaching a date and a number to every claim. For tokenisation that habit surfaces two figures. The first is Ethereum's Merge on September 15, 2026, which cut the network's energy use by roughly 99.95 percent, per the Ethereum Foundation. The second is that base-layer throughput remains around two dozen transactions per second.

The first figure says the environmental objection has lost its footing. The second says scaling was never solved at the base layer; it is solved on layer-twos, rollups and permissioned chains. For institutional tokenisation projects, scaling was never the real obstacle anyway, because they run on permissioned networks with a limited set of participants. The real obstacles are three: legal finality, accounting recognition, and the central bank's operating window.

The third obstacle gets the least attention. Conventional settlement runs inside fixed windows, T+1 or T+2, with end-of-day netting and overnight liquidity reconciliation. Blockchain runs 24 hours, 365 days. Twenty-four-hour settlement means end-of-day netting stops, which means every bank must hold more high-quality liquid assets outside its normal window. That extra buffer creates cost on the balance sheet. Part of the cost-saving promise in tokenisation marketing returns right there.

The second obstacle is quieter. If a tokenized treasury unit is not recognised by a regulator as a cash equivalent, its weight in the capital adequacy ratio rises, and a bank treasurer thinks twice before buying it. Between a successful pilot and a product that enters the balance sheet, there is a gap.

Much of the discussion still focuses on transaction speed. Anyone reading pilot results sees a different pattern: the technology test passes, the business test fails. The reason is simple. Technology teams run the pilot; treasury departments make the decision. The technology team shows a transaction settling in seconds. The treasurer asks which regulatory reporting window a 3am transaction falls into, and how it will be recognised on the balance sheet.

There is a bias here. We watch the technology because it is easy to measure, because it has timestamps, because it has benchmarks. The variable that actually makes the difference sits inside the balance sheet, where no scanner reaches it. A large part of Asia's tokenisation enthusiasm is still circling outside the settlement layer.

The most important question over the next eighteen months: which Asian jurisdiction will first give tokenized central bank money legal finality, not merely as a pilot but as a form usable in repo and interbank transactions? If no Asian central bank grants that recognition by mid-2026, the conclusion is that tokenisation never reached the core of Asian banking and stopped at the border. Fund sizes will keep growing. Balance sheets will not change.