Beyond the Digital Twin

In April, Khazanah sold RM100 million of sukuk that also lives on a blockchain. Malaysia's bond and sukuk market is worth RM2.256 trillion, so the pilot covered about four thousandths of one percent of it. The size was the first thing I noticed, because it tells you what the deal was for. Nobody moves a RM2.256 trillion debt market with RM100 million. A deal that size exists to prove the machinery works, and to leave behind a template the next issuer can copy without re-asking the same questions.

The template rewards a close read, because every choice in it is deliberate.

What stayed put

The sukuk's legal record never left RENTAS, the central bank's settlement system, where every ringgit bond has always lived. The blockchain carries a synchronised copy: a digital twin. Ownership cannot change on the chain alone. Every transfer also has to move through RENTAS, and issuance runs through FAST, the central bank's platform for issuing and tendering debt securities. The record a court recognises sits in those two systems, not on the ledger. For now, tokenisation here means running two systems in step for one instrument. RAM rated the credit impact neutral, which follows: the asset itself never moved.

The caution has a logic. Legal title is the last thing you move onto new rails, not the first. Prove custody works, then prove trading works, and only then move the record itself. HSBC's Orion, Euroclear and the IFC all began with the same twin structure. The pilot answered the custody question and left the harder questions of demand, liquidity and price for later.

Malaysia has been here before

Twice, in fact, both times offshore in Labuan. In November 2020, China Construction Bank's Labuan branch came within days of listing a digital bond sale of up to US$3 billion on Fusang, the Labuan digital exchange. The bank pulled it before listing, under scrutiny at home over the crypto association. A sale that size, on Malaysian rails, six years ago.

Then in October 2023, Fusang did list one: the world's first exchange-listed tokenised sukuk backed by a sovereign-linked instrument, built on a sukuk issued by the IILM. Note the words built on. Fusang wrapped the sukuk in a depository receipt and tokenised the receipt. Investors held a claim on a custodied instrument, one legal layer away from the sukuk itself. It is the same move global markets have used for a century to trade foreign shares without moving foreign title.

So Malaysia has now tried both known ways around the hard problem. Labuan tokenised a wrapper; Khazanah tokenised a mirror. In both, the legal title sat untouched in a conventional system. Nobody here, and almost nobody anywhere, has yet made the token the instrument. Moving the title itself is the frontier.

The Labuan deals teach one more thing. Rails and world-firsts were not enough: one deal died before listing, the other worked and stayed niche. A tokenised market needs the onshore system behind it: the regulator, the central bank, the domestic issuers. That is what makes 2026 different. This deal sits in the domestic market, inside RENTAS, and the Securities Commission co-sponsored it.

What a basis point is worth

Malaysia runs RM2.256 trillion in bonds and sukuk, about 36% of all the sukuk in the world. Shave one basis point off the annual cost of carrying that stock and you free roughly RM225 million a year. The RM225 million is arithmetic, on paper. Whether the basis points actually exist is a separate question. Here is where they sit:

• Margin. Settlement risk is priced in collateral: the longer a trade takes to settle, the more cash both sides park at the clearing house while they wait. When the US cut settlement by one day, about US$3 billion of clearing margin came back, roughly 23% of it. Tokenised settlement compresses the wait toward zero, and the collateral with it.
• Idle cash abroad. To pay for trades across borders, banks pre-fund accounts overseas and let the money sit waiting. Some US$24 trillion of wholesale payments cross borders each year, and moving it costs the corporates behind those payments more than US$120 billion a year, before counting the cost of all that parked liquidity. Cash that settles on-chain retires those balances.
• Hand-work. Count the parties a single coupon payment passes through today: issuer, paying agent, clearing house, global custodian, local sub-custodian, broker, investor. Five or six sets of books, each reconciled against the next, by people. On a token, the instrument pays its holders directly against one shared record. The saving recurs every payment date, across the entire stock.
• Trading. The issuance savings everyone advertises are still projections. On the US$8 billion of tokenised bonds issued worldwide so far, the BIS finds underwriting costs no cheaper than conventional deals; Cashlink, a German tokenisation platform, projects savings near 1.2% of a bond's value once volumes build, which is a projection, not a result. The measurable gain today is in trading: tokenised bonds change hands at a mean bid-ask spread of about 19 basis points, against 30 for comparable conventional bonds from the same issuers. Every trade pays that spread, so the saving compounds with turnover.

The stronger argument is access

Cost savings accrue to institutions. Access reaches everyone else. Government and high-grade sukuk are the safest ringgit assets there are, and today it takes an institution's ticket size to hold them. Tokenisation can break one institutional lot into small digital units inside the apps people already carry. The safest paper in the country stops being institutional-only. That is the outcome worth building a market for, and worth owning the rails of.

Who is building it

Watch CIMB. It arranged and tokenised the Khazanah sukuk. It is piloting a tokenised ringgit deposit with Bank Negara, the digital cash a tokenised bond would settle against. And in February it signed an early partnership with Ant International's Whale network for moving tokenised cash across borders. Asset, cash, distribution: a tokenised market runs on those three layers, and one bank is building all three. That is a bank taking the lead.

There is a fourth layer, and it is the one banks understand best: compliance. Banks are built on it. Knowing the customer, screening the counterparty, proving to a regulator that the wrong person cannot hold the wrong asset — that is the job. On shared rails it moves into the token itself: who may hold it, who may trade it, written into the instrument. Whoever's compliance the regulators trust becomes the gatekeeper of the market. For a bank, that layer is the moat.

What is still missing

Three pieces, none of them science fiction.

Shared rails. If every issuer runs its own private ledger, the market rebuilds the fragmentation it was trying to escape. The destination is shared public rails with the permissions built into the token. The world's biggest regulated names are already there: BlackRock runs BUIDL, a tokenised fund holding about US$2.5 billion, on public chains including Ethereum, and Franklin Templeton runs a US-registered mutual fund across eight public blockchains. On the bond side, the EIB issued €100 million on public Ethereum in 2021 and Siemens followed on Polygon in 2023 under Germany's Electronic Securities Act. The question left for Malaysia is not whether public rails can carry regulated assets. It is who signs off, and when.

On-chain ringgit. A tokenised bond needs tokenised cash to settle against, or the cash leg drags everything back to the old rails. Bank Negara is piloting tokenised deposits and a ringgit stablecoin, with policy due by end-2026. Cross-border, that cash has to plug into multi-bank networks, which is exactly what CIMB's Whale partnership is for.

Invisibility. This one is about people, and it gets skipped in every technical roadmap. Nobody adopts a market because the rails are clever. A treasurer wants settlement that just happens; a retail investor wants to tap an app they already have. No wallets, no seed phrases, no new vocabulary. The institutions that win this will be the ones whose customers never find out a blockchain was involved.

Four signals

• Bank Negara's end-2026 ruling on tokenised deposits and the ringgit stablecoin. The gate everything else waits behind.
• The twin going native: the token becoming the legal record, instead of a wrapper on it or a mirror of it.
• The first atomic settlement: a tokenised sukuk against tokenised cash, both legs in one step.
• The first public-rails issuance, with real retail access.

I expect most of these four inside two years. If they land, the RM100 million pilot will turn out to be where the market actually started.

Personal analysis on public data; not affiliated with the transaction.