Tokenized deposits vs stablecoins for treasury management: which should APAC businesses use in 2026?

For treasury teams across Asia-Pacific, the tokenized deposits vs stablecoins decision has moved from theory to budget line item. As of August 2026, both instruments are live at scale: stablecoins settle an estimated $10-15 trillion in annual on-chain transaction volume globally, while tokenized deposit programs run by banks including JPMorgan (Kinexys, formerly Onyx), HSBC, and several Singapore-based institutions now handle meaningful wholesale flows. The short answer is that they solve different problems. Stablecoins are bearer instruments issued by non-bank entities, useful for payments reach and 24/7 settlement outside the banking system. Tokenized deposits are digital claims on your existing bank account balance, useful when you want programmable money that stays inside the regulated banking perimeter. Most corporate treasurers in Asia-Pacific will end up using both, allocated by counterparty type, jurisdiction, and risk appetite — and the allocation logic is exactly where modern treasury intelligence tooling earns its keep.

What Tokenized Deposits Actually Are

Also worth reading: How does AI cash flow management work for multi-currency businesses across the Asia-Pacific region? · What are the leading ASEAN treasury automation trends reshaping corporate cash management in 2026? · Is it worth moving from Excel spreadsheets to a cloud TMS? What's the real ROI of cloud treasury management vs spreadsheets?

A tokenized deposit is a digital representation of a commercial bank deposit, recorded on a distributed ledger or permissioned blockchain rather than only in the bank's core ledger. Critically, you hold a claim against your own bank, denominated one-to-one in fiat currency, with the balance remaining on the bank's balance sheet and covered by whatever deposit insurance regime applies — up to S$100,000 per depositor per bank under Singapore's Deposit Insurance Scheme, or HK$800,000 in Hong Kong since October 2024. There is no separate reserve pool, no issuer bankruptcy risk distinct from your bank's own solvency, and no redemption mechanism to stress-test because redemption is simply a ledger entry reversal.

The practical consequence is that tokenized deposits inherit the credit profile of your existing bank relationship. If you hold deposits at a G-SIB or a well-capitalized regional bank, your tokenized balance carries essentially the same overnight risk as the account itself. This makes them attractive for intra-group cash pooling, supplier financing within a known banking network, and any flow where both counterparties bank with the same institution or its partners. JPMorgan's Kinexys Digital Payments has processed over $1.5 trillion in notional volume cumulatively, largely from exactly this use case: corporate clients moving value between their own accounts instantly, outside RTGS operating hours.

The limitation is equally clear. Tokenized deposits do not travel well between banking ecosystems. A deposit token at Bank A is, by design, worthless to a counterparty who refuses to open an account at Bank A. Interoperability projects — including work coordinated through the BIS's Project Agorá and various interbank token networks — aim to fix this, but as of mid-2026 cross-bank tokenized deposit settlement remains patchy, concentrated in a handful of corridors such as USD-SGD and EUR-GBP wholesale rails.

What Stablecoins Are and Where They Fit

Stablecoins are blockchain-native tokens pegged to a reference asset, most commonly the US dollar, backed by reserve portfolios of short-dated US Treasuries, reverse repos, and cash at regulated custodians. The two dominant issuers, Tether (USDT) and Circle (USDC), together represent roughly 85-90% of total stablecoin market capitalization, which stood near $250 billion in early 2026 after growing roughly 40% year-on-year. Circle's USDC reserves are held primarily in the Circle Reserve Fund, a government money market fund managed by BlackRock, with daily transparency reporting; Tether publishes quarterly attestations rather than full audits, which remains a legitimate point of criticism for conservative treasurers.

The regulatory environment has firmed considerably. In the United States, the GENIUS Act signed in July 2025 established federal licensing for payment stablecoin issuers, requiring 100% reserves in high-quality liquid assets, monthly disclosure, and prohibitions on paying interest to holders. In Hong Kong, the Stablecoins Ordinance took effect on 1 August 2025, requiring issuers of HKD-referenced stablecoins to be licensed by the HKMA with fully backed reserves. Singapore's MAS finalized its stablecoin regulatory framework in 2023, applying to single-currency stablecoins pegged to SGD or G10 currencies with circulation above S$5 million. Japan's Payment Services Act amendments have permitted bank-issued stablecoins since 2023, with MUFG's Progmat platform enabling yen-denominated issuance.

For treasurers, stablecoins offer what tokenized deposits cannot: portability. A USDC balance moves between any wallet, exchange, or counterparty in seconds, settles on weekends and holidays, and reaches counterparties in jurisdictions where correspondent banking has thinned out. That last point matters enormously in Asia-Pacific, where de-risking has reduced correspondent relationships in several markets and where intra-ASEAN trade settlement increasingly looks for alternatives to slow USD corridors.

Head-to-Head Comparison

Neither instrument dominates. The comparison table shows why sophisticated treasuries treat them as complements in a liquidity stack rather than competitors for a single mandate.

Why Treasury Teams Are Moving Now

Three forces converged between 2024 and 2026. First, settlement economics: traditional cross-border B2B payments through correspondent chains still take one to three business days and cost an average of 1.5% to 3% all-in once FX spreads, intermediary fees, and investigation costs are counted. Stablecoin settlement compresses this to minutes at a marginal cost often below 0.1%, and tokenized deposit networks deliver similar speed within their footprints. For a company moving $50 million monthly across borders, the difference between 2% and 0.2% friction is $1.08 million annually — a number that gets CFO attention regardless of enthusiasm for the underlying technology.

Second, working capital timing. Corporate receivables that settle on T+2 trap cash that could offset revolver draws or earn money-market returns. At 4-5% short-term USD rates through much of 2025-2026, every day of trapped float on a $100 million receivables book costs roughly $13,000. Instant settlement converts that float into deployable liquidity, and AI-driven cash-flow forecasting platforms can quantify the benefit per corridor before you commit.

Third, regulatory clarity removed the compliance excuse. Before the GENIUS Act and Hong Kong's ordinance, holding stablecoins on a corporate balance sheet raised accounting and audit questions that many controllers refused to touch. With licensed issuers, defined reserve requirements, and clearer treatment under frameworks like IFRS 9 (stablecoins generally classified as financial assets at amortized cost or fair value through P&L depending on business model), the audit conversation is now tractable. Banks themselves have shifted from resistance to participation: Citigroup analysts projected the stablecoin market could reach $1.6 trillion to $3.7 trillion by 2030, and Bain's 2026 wholesale banking research describes stablecoin infrastructure as rewiring correspondent banking economics rather than merely adding a rail.

Practical Steps for an APAC Treasury

Start with a corridor and use-case inventory. Map your top ten payment corridors by volume, current settlement time, and all-in cost. Identify which flows are intra-group (strong candidates for tokenized deposits if your banks operate programs), which are vendor payouts to crypto-native or thin-banking counterparties (stablecoin candidates), and which genuinely need neither because local instant-payment systems like PayNow, PromptPay, UPI, or DuitNow already solve them cheaply. A common error is deploying blockchain rails where a domestic real-time payment system would be simpler and cheaper.

Next, establish custody and policy before making a first trade. Decide between direct self-custody with hardware wallets and multi-signature controls, qualified custodians such as Fireblocks, Anchorage Digital, or Hex Trust (which operates in Hong Kong and Singapore), or bank-hosted solutions. Write a formal digital-asset treasury policy covering counterparty limits, approved chains (Ethereum mainnet, Base, Solana, and various permissioned networks each carry different fee and finality profiles), reconciliation procedures, and incident response. Boards and auditors will ask for this document; having it drafted before your first transaction signals control maturity.

Then pilot small. A typical sequence runs: internal transfer between two of your own wallets (week one), a single trusted vendor payout of $10,000-$50,000 (weeks two to four), then scaling to a defined percentage of a chosen corridor — say 10% of monthly volume — while comparing realized costs against the legacy rail. Track four metrics: end-to-end settlement time, all-in cost per transaction, exception rate, and reconciliation effort in staff hours. Most teams find reconciliation, not settlement, becomes the operational bottleneck, which argues for ERP-integrated tooling rather than manual wallet exports.

Finally, negotiate with your banks. Several APAC banks now offer tokenized deposit pilots to corporate clients on request, and asking directly accelerates access. If your lead bank has no program, ask whether they participate in any interbank token network; if not, that is useful information about where your bank sits on the adoption curve.

Common Mistakes and How to Avoid Them

The most expensive mistake is treating stablecoin holdings as risk-free cash equivalents. Even fully reserved stablecoins carry depeg risk during stress: USDC traded as low as $0.87 in March 2023 when Silicon Valley Bank failed and $3.3 billion of its reserves were briefly stranded there. The peg recovered within days, but a treasurer holding payroll funds through that weekend learned the lesson viscerally. Mitigate by capping single-issuer exposure, monitoring reserve disclosures, and converting to banked fiat on a defined cadence rather than accumulating large balances.

The second mistake is ignoring chain and venue risk. Sending USDC on the wrong network, interacting with a spoofed contract address, or leaving funds on an exchange that later freezes withdrawals are operational failures, not technology failures. FTX's collapse in November 2022 — where customer stablecoins were moved off-platform and exchanged — remains the canonical case study for why custody discipline matters more than issuer choice. Use allowlisted addresses, test transactions, and segregation between trading and settlement wallets.

Third, teams underestimate tax and accounting complexity. In several APAC jurisdictions, converting fiat to stablecoin and back can trigger taxable events or at minimum requires documented valuation methodology. Engage tax advisors before scaling; retrofitting records across hundreds of on-chain transactions is far costlier than designing the data capture upfront.

Fourth, some organizations chase tokenized deposits expecting public-chain-style openness and are disappointed by permissioned-network constraints, while others adopt stablecoins expecting bank-grade recourse and discover that consumer-protection style dispute mechanisms barely exist. Set expectations per instrument according to the table above, not according to marketing language from either camp.

Costs, Pricing, and Economics

Direct costs are modest but nonzero. Stablecoin transfers on Ethereum mainnet can cost $1-10 in gas during congestion; on Layer-2 networks like Base or Arbitrum, or on Solana, fees typically run under $0.05. Acquiring stablecoins through licensed OTC desks or exchanges involves spreads of roughly 5 to 25 basis points on institutional size, plus possible platform fees. Qualified custody runs roughly 10 to 50 basis points annually on assets under custody depending on features. Tokenized deposits, by contrast, usually carry no explicit token fee — costs hide in banking relationship pricing, integration project budgets (commonly $50,000-$500,000 for ERP and API integration at mid-size corporates), and internal engineering time.

Compare these against the baseline: SWIFT-correspondent cross-border payments averaging 1.5-3% all-in, and wire fees of $15-50 per transaction plus unfavorable retail FX spreads. The break-even math favors digital rails for any organization settling more than roughly $1 million monthly across borders, though the exact threshold depends on your negotiated banking terms. Note also the opportunity cost angle: idle stablecoin balances earn nothing under the GENIUS Act's interest prohibition, so large structural balances belong in T-bill ladders or money market funds, with stablecoins used as a transit medium rather than a store of value.

When to Act, and When to Wait

Act now if three conditions hold: you move material cross-border volume through slow or expensive corridors, you operate in or serve markets with degraded correspondent banking access, and your finance leadership will fund proper custody and policy infrastructure. Companies meeting all three are leaving measurable money on the table every month they wait, and competitor adoption in trade-heavy sectors — electronics manufacturing, commodities, logistics — accelerated visibly through 2025.

Wait, or proceed cautiously, if your flows are predominantly domestic (local instant payment systems already suffice), if your counterparties show zero interest in digital settlement (a rail needs two sides), or if your finance function lacks capacity to build controls properly — a half-implemented program creates more risk than the status quo. Also watch the interoperability question: if BIS Project Agorá and successor initiatives deliver genuine cross-bank tokenized deposit settlement by 2027-2028, the calculus may shift toward bank tokens for flows currently served by stablecoins, particularly for treasurers uncomfortable with non-bank issuer exposure.

The realistic 2026 posture for an APAC operator is hybrid: tokenized deposits for intra-group and same-bank settlement where they are available, regulated stablecoins for external cross-border payments and vendor reach, traditional rails retained as fallback, and a forecasting layer that models liquidity across all three in real time. That last piece — continuous, AI-assisted visibility into where cash sits, in what form, and what it costs to move — is the actual competitive edge, because the instruments will keep changing while the discipline of measuring them does not.