Global Rails: 24/7 Tokenised Settlement, Venue Pluralism and the De-Americanisation of Retail Financial Onboarding by 2031openHighorigin: briefs
Output
| Thesis probability | 80.3%from a 95.7% prior · 18 active evidence items |
| Reasonable range | 55% – 99% tick = prior · marker = posterior |
| Historical analogy strength | 72/ 100 · 21 analogues in the reference class |
| Current evidence strength | 99/ 100 · 18 active items |
| Evidence quality | 70/ 100 |
| Contradictory evidence | High |
| Confidence in probability | High |
| Next major confirmation signal | Two hard, falsifiable prints: (1) a top-five broker/exchange publishing tokenised-equity volume with named market makers and quoted spreads, and (2) the first CCP, prime broker or insurance regulator formally accepting a tokenised money market fund as eligible margin/collateral. Secondary: the next Artemis/rwa.xyz adjusted (non-crypto-native) stablecoin settlement print — needs to be compounding above ~80% y/y to reach $1trn by 2031 — and any official SEBI/CSRC/CBN move on outbound retail investing. A single large de-peg or tokenised-custody failure is the main downside trigger to watch. |
Posterior log-odds = prior log-odds (reference-class base rate) + sum of quality x independence weighted log likelihood ratios. Research-informed estimate until enough theses resolve for Brier calibration.
Formalized thesis
By 31 December 2031, retail and institutional investors routinely operate across 24/7, multi-currency, multi-venue rails: tokenised equities and stablecoin settlement exceed $1trn in annual on-chain volume, at least three credible non-NYSE/Nasdaq or non-US listing/payment venues capture measurable primary-listing or cross-border payment share, and a majority of new retail financial relationships globally are opened on non-US platforms.
Source observations
- 2026-09-14Financial market structure is fragmenting along political and geographic lines: Texas Stock Exchange winning its first listing, Kazakhstan-China QR payments extending the yuan, Vale eyeing a yuan bond investors will increasingly need multi-venue, multi-currency access.
- 2026-09-17Money becomes programmable and always-on: tokenised equities, stablecoin rails and 24/7 settlement shift the unit of financial trust from the broker's opening bell to the chain. Over years this erodes the distinction between 'investing' and 'transacting'.
- 2026-09-25Mobility and payments are globalising around Asian rails — Tencent launching a payments app for 'China-maxxing' foreign tourists, Indian retail brokerage democratization — the next billion financial relationships form outside US platforms.
Causal mechanism
Access and optionality: savers in fast-growing, capital-scarce, inflation-prone economies want dollar exposure, global equities and instant cross-border transfer without a US bank relationship; institutions want capital efficiency (intraday/atomic settlement releases collateral), and everyone wants their money to behave like the rest of their phone — continuous, instant, borderless. A secondary motivation is jurisdictional diversification: after 2022 sanctions and 2025-26 tariff/rates volatility, holding claims solely inside the US stack is perceived as a political risk.
T+1/T+2 settlement lag, banking hours and weekend closures, correspondent-banking nostro/vostro chains (3-5 intermediaries, 2-6% cost on remittances), FX cut-offs, minimum-ticket and accredited-investor gating, and the geographic requirement to have a US brokerage relationship to hold US equities. Fractionalisation plus programmable settlement removes both the capital minimum and the calendar.
Public/permissioned blockchains with institutional custody (Fireblocks, Anchorage, BNY/State Street digital custody), regulated stablecoins under MiCA/GENIUS-style regimes, tokenised money market funds (BUIDL/BENJI-class), the SEC's five-year innovation exemption for tokenised stock trading, atomic DvP/PvP, UPI/Pix-style instant domestic rails with cross-border linkages (Project Nexus, mBridge), and AI agents that make multi-venue, multi-currency complexity invisible to the user.
Issuers: cheaper, faster, 24/7 capital raising and access to a global retail bid without ADR/depositary friction. Intermediaries: float on stablecoin reserves at ~4-5% policy rates is a large, near-costless revenue line; brokers monetise 24/7 order flow and weekend spreads. Users: 50-150bps saved on FX and remittances, yield on cash balances, and access to assets previously gated. Venues: listing fees and data revenue are the prize non-US exchanges are competing for.
Users must believe the token is a legally enforceable claim on the underlying share/dollar, that the custodian/issuer will redeem at par under stress, and that a foreign regulator will protect them. This is trust infrastructure that must precede convenience: reserve attestations, bankruptcy-remoteness, insurance, deposit-style protection, and at least one credible stress event survived without loss. A single large stablecoin or tokenised-custody failure resets the clock 3-5 years.
Mobile-first super-apps and neobrokers (Zerodha/Groww, Nubank, Revolut, Robinhood international, Binance/Coinbase international, Interactive Brokers white-label), embedded finance inside messaging and e-commerce platforms, plus incumbent distribution defensively adding tokenised products (BlackRock, Fidelity, Schwab, Nasdaq/DTCC tokenised collateral pilots).
Strong and two-sided: liquidity begets liquidity in trading venues; stablecoins exhibit classic money network effects (USDT/USDC dominance is self-reinforcing across exchanges, remittance corridors and merchant acceptance); a listing venue needs both issuers and an investor base. Counterweight: fragmentation across chains and jurisdictions can stall the flywheel unless interoperability/bridging is solved, and the incumbent NYSE/Nasdaq/DTCC network effect is one of the strongest in global finance.
Low for new retail accounts (the population that matters for condition (c)) — an app download and KYC. High for institutions: legacy custody, accounting, tax lots, risk systems, collateral eligibility rules and fiduciary/prudential capital treatment. Corporate issuers face near-prohibitive switching costs for existing US listings, so venue share must come from NEW listings and from companies that would previously have chosen the US.
The binding variable. Positive: SEC innovation exemption (2026), MiCA, GENIUS-type stablecoin statutes, Hong Kong/Singapore/UAE/UK digital-securities regimes, Project Nexus. Negative: securities-law extraterritoriality, KYC/AML and travel-rule friction, capital controls (India, China, Nigeria), Basel crypto-asset capital charges, bank lobbying against deposit disintermediation, and the possibility that US authorities restrict non-US platforms offering US-equity tokens to their citizens. Conditions (b) and (c) are as much geopolitical as technological.
Historical analogues
| Analogue | Period | Mechanism | Similarity | Occurred | Yrs to mainstream | Lesson | |
|---|---|---|---|---|---|---|---|
Tether/USDC dollarisation in emerging markets | 2019-2026 | A privately issued, 24/7, bearer-style dollar claim substituted for correspondent banking in Argentina, Turkey, Nigeria and Vietnam; users accepted issuer credit risk in exchange for access to dollars. | 88 | ✓yes | 6 | Demand for offshore dollar access is the strongest observed driver; stablecoin settled value already runs in the trillions annualised on gross measures — the measurement definition, not adoption, decides whether the $1trn test is passed. | |
Alipay/WeChat Pay displacing cards in China | 2013-2018 | Super-app distribution wrapped custody, payments, credit and investing (Yu'e Bao) behind one interface; hundreds of millions of first financial relationships opened on non-bank, non-US platforms. | 86 | ✓yes | 5 | The 'majority of new retail relationships are non-US' condition has already been true in China for a decade — the arithmetic of condition (c) may be won by demographics rather than by tokenisation. | |
M-Pesa mobile money in Kenya | 2007-2013 | Non-bank rails delivered by a telco to an under-served population; trust in agents and float substituted for trust in banks; leapfrogging of the incumbent stack entirely. | 84 | ✓yes | 5 | Leapfrogging happens where the incumbent option is absent or expensive, not where it is merely inconvenient — supports condition (c) in EM, not in the US/EU. | |
Pix in Brazil | 2020-2024 | Central-bank mandated instant rail; ~150m users in four years, displacing cards and cash for P2P and increasingly merchant payments, and pulling retail into Nubank-class non-US platforms. | 84 | ✓yes | 3 | Mandated interoperability plus free pricing collapses adoption time to 2-4 years. Where a regulator mandates it, the horizon assumption is conservative; where it must be voluntary, double the timeline. | |
UPI in India | 2016-2023 | State-sponsored instant, interoperable, zero-MDR payment rail; volume went from zero to >10bn transactions/month, and India began exporting the rail to other jurisdictions. | 83 | ✓yes | 6 | Public-infrastructure rails scale faster than private ones and can become credible cross-border venues (UPI-PayNow, Nexus) — the most likely path for three non-US payment venues taking measurable share. | |
Binance and offshore 24/7 crypto exchanges capturing global retail | 2017-2022 | A non-US, always-on, multi-asset, multi-currency venue captured the majority of global retail derivative and spot flow by removing KYC/geography/hours friction — until regulators forced onshore compliance. | 81 | ✓yes | 4 | Offshore venues scale fastest precisely where trust infrastructure is weakest — and then face a compliance reckoning (FTX, the 2023 DOJ settlement) that transfers share back to regulated players. Expect one such reckoning inside the forecast window. | |
Robinhood / zero-commission mobile brokerage and global copycats | 2015-2021 | Mobile-first, fractionalised, gamified access removed minimum-ticket and paperwork friction, bringing tens of millions of first-time investors in, largely outside incumbent wirehouses. | 78 | ✓yes | 5 | Fractionalisation plus a smartphone is the single most powerful driver of NEW account formation — and the cohort is young, global and indifferent to venue brand. | |
DTC/dematerialisation after the paperwork crisis | 1968-1983 | Physical certificates replaced by book-entry claims — the original tokenisation. Investors gave up direct possession (control) for settlement speed and cost (access). | 76 | ✓yes | 12 | Settlement-layer transformations take 10-15 years and are driven by an operational crisis plus a regulatory mandate, not by user demand. The incumbent CSD captured the value, not a challenger. | |
Online discount brokerage and retail direct market access | 1996-2007 | Interface abstraction plus commission collapse moved custody and execution from full-service brokers to self-directed screens; users traded advice (control by proxy) for access and price. | 74 | ✓yes | 8 | Retail adopts a new venue for cost and access but only after a bull market supplies the motivation; adoption stalls hard in drawdowns. | |
Eurodollar market — offshore dollar deposits escape onshore rails | 1957-1975 | A regulated, calendar-bound domestic system (Reg Q ceilings, reserve requirements) pushed dollar intermediation offshore; a parallel venue captured enormous share of the SAME asset without displacing the domestic currency. Control traded for access. | 71 | ✓yes | 12 | Offshore venues win share of a domestic asset when domestic regulation imposes a cost, not when technology is better. The dollar strengthened even as its plumbing de-Americanised — condition (b) and (c) can be satisfied while US assets remain dominant. | |
NASDAQ's electronic quotation system displacing floor-based OTC trading | 1971-1990 | Complexity (fragmented dealer quotes) hidden behind an electronic interface; a new venue with no floor captured listings the incumbent considered unlistable. | 70 | ✓yes | 15 | New venues capture the marginal (new, smaller, faster-growing) issuer first and only later contest the incumbent's franchise — 19 years to become a credible peer. | |
CLS Bank and continuous linked (PvP) FX settlement | 1997-2008 | Atomic settlement eliminated Herstatt risk; a new global utility captured a majority of the world's FX settlement value within ~6 years of launch. | 69 | ✓yes | 7 | Institutional settlement value can scale to trillions quickly once regulators bless a single mechanism — evidence that the $1trn on-chain threshold is modest if bank consortia adopt. | |
Reg ATS / ECNs / Reg NMS fragmenting US equity trading | 1997-2007 | Regulatory permission (Order Handling Rules, Reg ATS) created a legal on-ramp for new venues; Island/Archipelago/BATS took ~30-40% of volume within a decade before being absorbed. | 68 | ✓yes | 8 | A regulatory exemption is the trigger; venue share shifts materially within 5-10 years, but incumbents recapture economics through acquisition and data revenue. Venue pluralism ≠ incumbent profit loss. | |
Eurex capturing the Bund futures contract from LIFFE | 1997-1999 | An electronic, 24-hour-capable venue took ~100% of an incumbent's flagship contract in roughly 18 months once fee and access frictions were removed. | 68 | ✓yes | 2 | When the product is fungible and the technology strictly dominates, venue migration can be near-total and extremely fast — the tail scenario for tokenised venues is not gradual. | |
Index/ETF wrapper displacing direct security ownership | 1993-2010 | Investors traded direct control of holdings for cheap, liquid, intermediated access; the wrapper became mandatory infrastructure and the sponsors captured the pricing power. | 68 | ✓yes | 14 | When control is traded for access, the value accrues to whoever operates the wrapper and the custody layer — not to the exchange and not to the end investor. | |
SWIFT replacing telex for interbank messaging | 1973-1987 | Trust infrastructure (a member-owned, standardised, audited messaging network) built before convenience was possible; adoption required collective action among rivals. | 67 | ✓yes | 10 | Shared financial infrastructure requires a neutral governance body; without one, competing tokenisation consortia fragment and the $1trn threshold slips right. | |
Eurobond market and the migration of dollar issuance to London | 1963-1980 | A tax (the US Interest Equalization Tax) made the incumbent listing/issuance venue expensive; issuers moved to a foreign venue with lighter disclosure, creating a permanent non-US primary market for dollar claims. | 66 | ✓yes | 8 | Primary-listing share moves fast when a policy wedge appears; it rarely reverses even after the policy is repealed. Watch for a policy wedge, not for investor preference. | |
MiFID I and the rise of Chi-X/BATS Europe | 2007-2014 | Regulation abolished concentration rules; pan-European MTFs captured 30%+ of lit trading in national blue chips within 5 years, fragmenting national exchange monopolies. | 66 | ✓yes | 5 | Trading share fragments quickly; LISTING share almost never does. Distinguish the easy leg (secondary volume) from the hard leg (primary listings) in condition (b). | |
CME Globex and the normalisation of near-24-hour futures trading | 1992-2005 | Continuous electronic sessions replaced pit hours; the market discovered that demand for off-hours risk transfer existed once the interface made it costless. | 66 | ✓yes | 10 | 24/7 access is adopted but liquidity stays concentrated in the legacy session for a decade-plus. 'Routinely operate 24/7' can be true on access while false on volume distribution. | |
Chinese issuer listing migration from NYSE/Nasdaq to HKEX and STAR Market | 2019-2024 | Geopolitical and audit-regulatory wedge (HFCAA) pushed primary listings to non-US venues; HKEX and STAR each took meaningful global IPO-proceeds share. | 54 | ✓yes | 4 | Non-US listing share is won geopolitically and in bursts; two or three such wedges (India, Gulf, Hong Kong) are sufficient to satisfy condition (b) without any tokenisation at all. | |
Gulf exchanges (Tadawul, ADX, DFM) capturing global IPO proceeds | 2019-2025 | Sovereign privatisation pipelines plus local liquidity pools created credible non-US primary venues; Saudi Aramco/ADNOC-class listings pushed the region into top-5 IPO league tables. | 50 | ✓yes | 5 | Primary-listing share follows domestic capital pools and state agendas; this is an already-in-progress trend that de-risks one leg of the thesis. |
Reference class: regulator-enabled financial market-structure migrations in which (i) settlement or access friction is removed by a new technical/legal layer, (ii) users delegate custody or control to an intermediary/ledger in exchange for continuous, cheaper, borderless access, and (iii) share shifts toward venues outside the previously dominant jurisdiction. The class deliberately mixes three sub-classes because the thesis is a conjunction of three distinct bets: (A) settlement-layer transformations (DTC dematerialisation, SWIFT, CLS, stablecoins) which are slow, trust-gated, 7-15 years, but reliably reach trillion-dollar scale once a regulator blesses them; (B) venue-share migrations (ECNs, MiFID/Chi-X, Eurex/LIFFE, NASDAQ, Eurodollar/Eurobond, HKEX/Gulf listings) which are fast for secondary volume (2-8 years) and slow-to-never for primary listings, and are driven by policy wedges rather than user preference; and (C) retail onboarding leapfrogs (M-Pesa, UPI, Pix, Alipay, Robinhood, Binance) which are the fastest in the class (3-6 years) and are demographically determined — the marginal new account holder is already non-US. Base rate of the pattern occurring at all in this class is very high (21/21 in the retrieved set, which itself signals survivorship bias); the informative variance is in TIMING and in the precise thresholds, not in direction. Median years-to-mainstream across the set is ~6-7, with sub-class A at ~10-12 and sub-class C at ~4-5. Measured from the 2026 SEC innovation exemption and MiCA/GENIUS implementation as t=0, the 2031 horizon sits at year 5 — at or slightly inside the median for the fastest sub-classes and well inside the median for settlement-layer change.
Caveats: 1) The thesis is a three-part CONJUNCTION; even if each leg is individually more likely than not, the joint probability is materially lower and the system should score legs separately before combining. Leg (c) is close to an arithmetic near-certainty (new account formation is already dominated by India, Brazil, Indonesia, Nigeria and China), leg (b) is well advanced (Gulf + HKEX + India already near or above 1% listing/payment share), and leg (a) is the genuinely contested one and depends almost entirely on measurement definition — gross stablecoin transfer volume already exceeds $1trn on some adjusted methodologies, while tokenised EQUITY settlement is under $100bn. The thesis could be 'trivially true' or 'clearly false' depending on which denominator the resolver picks; the formalisation must be pinned to a named source before scoring. 2) Selection bias: historical analogues are drawn from transitions that succeeded and are therefore remembered; failed parallel rails (e.g. 1990s digital cash, DTCC blockchain repo pilots abandoned 2018-22, Libra/Diem, numerous failed regional exchanges, mBridge's stalled BIS exit) are under-represented. Libra in particular is the cleanest counter-case: overwhelming distribution, superior technology, killed by regulators in 18 months. 3) The analogue similarity scores conflate 'trading venue' and 'listing venue' dynamics; MiFID and ECN evidence supports volume fragmentation but is actively EVIDENCE AGAINST primary-listing migration. 4) Several analogues (UPI, Pix, Alipay) were state-mandated, which is not available to tokenised equities in most jurisdictions — their speed should be discounted by roughly half when transferred. 5) Correlated failure mode: a single large stablecoin depeg, tokenised-custody bankruptcy, or a US move to block foreign platforms from offering US-equity tokens would simultaneously damage all three legs, so the legs are NOT independent on the downside. 6) Macro regime dependency not present in most analogues: a ~5% risk-free rate makes stablecoin float highly profitable (accelerant) but also makes speculative retail account formation cyclical (decelerant); the 1996-2007 brokerage analogue shows retail adoption stalls hard in drawdowns. 7) Incumbent response is systematically underweighted in the analogue set: NYSE/Nasdaq/DTCC extending to 24/5, launching their own tokenised collateral, or acquiring challengers (the ECN endgame) would make the thesis directionally right and the venue-share leg wrong.
Evidence
| # | Claim | Direction | Strength | Quality | Indep. | Weight | LR | Wtd log-LR | Kind | Source | Date |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 0 | The SEC granted a five-year innovation exemption for tokenised stock trading in September 2026 after the Clarity Act stalled, explicitly opening a legal path to 24/7 tokenised equity venues and re-rating brokers, custodians and stablecoin rails. | supports | strong | 0.85 | 0.90 | 0.77 | 3.00 | +0.840 | Observed fact | Daily brief 2026-09-17 (regulatory event) | 2026-09-17 |
| 1 | Stablecoin float is a near-costless revenue line at ~5% policy rates (10y ~5.15%, 30y 5.44% post-2004 highs), giving issuers and distributing brokers a structural profit incentive to push settlement on-chain; the same rate level makes bank sweep/deposit margins the thing being disintermediated. | supports | strong | 0.80 | 0.75 | 0.60 | 3.00 | +0.659 | Observed fact | Rates data in briefs 2026-09-24/25 + stablecoin reserve economics | 2026-09-25 |
| 2 | Condition (b) is close to already satisfied on the listing leg: HKEX, the Indian exchanges, the Gulf bourses (Tadawul/ADX) and Shanghai/STAR have each repeatedly cleared ~1%+ of global IPO proceeds in recent years, and Chinese issuer migration from Nasdaq to HKEX/STAR is a completed structural shift. | supports | strong | 0.70 | 0.85 | 0.59 | 3.00 | +0.654 | AI inference | Dealogic/LSEG league-table pattern; analogue set | 2026-09-25 |
| 3 | Condition (c) is arguably already true on raw counts: India alone adds tens of millions of demat/brokerage accounts annually (Zerodha/Groww), plus Nubank, Revolut, Indonesian/Nigerian/Gulf wallets — non-US-headquartered platforms almost certainly already open the majority of the world's new retail financial relationships; the open question is measurement and reporting, not behaviour. | supports | strong | 0.65 | 0.80 | 0.52 | 3.00 | +0.571 | AI inference | National regulator counts (SEBI/NSE), company disclosures; brief reference to Indian retail brokerage | 2026-09-25 |
| 4 | Attention data shows consumers actively delegating complexity to agents ('ai agent builder', 'ai phone assistant' among fastest-rising queries, Edge AI +25.6, AI Software & Agents +28.5) — the interface layer that makes multi-venue/multi-currency settlement invisible is being adopted now. | supports | moderate | 0.60 | 0.70 | 0.42 | 2.00 | +0.291 | Observed fact | Attention movers, 2026-09-25 | 2026-09-25 |
| 5 | Instant domestic rails (UPI, Pix) have already achieved majority-share payment behaviour in their home markets and are being linked cross-border (Project Nexus, mBridge), demonstrating that the population in question switches rails within 3-5 years when cost and latency improve by an order of magnitude. | supports | strong | 0.85 | 0.90 | 0.77 | 3.00 | +0.840 | Observed fact | BIS/CPMI, NPCI, BCB data | 2026 |
| 6 | Jurisdictional-diversification demand is rising: European institutions repatriating bullion from the US, $365trn global debt, sanctions memory and 2025-26 tariff volatility all push savers and treasurers toward holding claims outside a single legal stack. | supports | moderate | 0.50 | 0.60 | 0.30 | 2.00 | +0.208 | AI inference | Brief 2026-09-13/25 commentary | 2026-09-25 |
| 7 | Tokenised equity settlement volume remains de minimis — low single-digit billions annually versus ~$1trn+ needed in combination with stablecoins — and virtually all current on-chain value is stablecoin transfer, most of which is crypto-native trading flow explicitly excluded by the metric. | contradicts | strong | 0.80 | 0.90 | 0.72 | 0.33 | -0.791 | Observed fact | rwa.xyz / Artemis adjusted-volume methodology | 2026 |
| 8 | Institutional switching costs are high and unresolved: collateral eligibility, Basel crypto capital charges, fiduciary rules, accounting/tax-lot systems and prudential treatment mean bank and asset-manager settlement migration lags retail by years even where the technology works. | contradicts | strong | 0.75 | 0.80 | 0.60 | 0.33 | -0.659 | Observed fact | Basel Committee crypto-asset standard; custody practice | 2026 |
| 9 | Capital controls in the largest account-growth markets (India, China, Nigeria) directly obstruct the cross-border multi-currency leg of the thesis, and regulators there have repeatedly acted against offshore crypto/brokerage access. | contradicts | strong | 0.80 | 0.85 | 0.68 | 0.33 | -0.747 | Observed fact | RBI/SEBI, CSRC, CBN policy record | 2026 |
| 10 | Trust infrastructure has not survived a clean stress test: exchange hacks continue to be reported in 2026, and a single large stablecoin de-peg or tokenised-custody failure would reset institutional adoption 3-5 years. | contradicts | moderate | 0.60 | 0.70 | 0.42 | 0.50 | -0.291 | AI inference | Brief 2026-09-25 (exchange hacks); mechanism trust requirement | 2026-09-25 |
| 11 | The NYSE/Nasdaq/DTCC network effect remains among the strongest in finance: existing issuers face near-prohibitive relisting costs, US equity depth is at record highs (Nasdaq printing new highs), and a hawkish Fed plus a firm dollar is currently pulling global capital toward, not away from, US assets. | contradicts | strong | 0.85 | 0.80 | 0.68 | 0.33 | -0.747 | Observed fact | Brief 2026-09-25 market data | 2026-09-25 |
| 12 | US authorities retain the option to restrict non-US platforms offering tokenised US equities to US or even to foreign persons via extraterritorial securities enforcement; the five-year exemption is a domestic carve-out, not a global liberalisation, and Congress failed to pass the Clarity Act. | contradicts | moderate | 0.70 | 0.75 | 0.52 | 0.50 | -0.364 | AI inference | Brief 2026-09-17; US securities-law precedent | 2026-09-17 |
| 13 | Bank lobbying against deposit disintermediation intensifies as stablecoin yield-bearing substitutes compete with sweep and savings balances at 5% rates; GENIUS-type statutes deliberately restrict interest-bearing stablecoins, capping the main retail adoption incentive. | contradicts | moderate | 0.65 | 0.70 | 0.45 | 0.50 | -0.315 | Observed fact | GENIUS Act structure; bank trade-association positions | 2026 |
| 14 | Incumbents are defensively adopting rather than being displaced: DTCC/Nasdaq tokenised-collateral pilots, BlackRock/Fidelity tokenised money funds and extended US trading hours mean the technology diffuses inside the existing stack, satisfying condition (a) while blunting conditions (b) and (c). | contradicts | moderate | 0.70 | 0.60 | 0.42 | 0.50 | -0.291 | Consensus expectation | Industry pilots; brief 2026-09-17 | 2026 |
| 15 | Roughly half of tokenised stock trading volume under the new US regime is occurring outside US market hours, indicating the continuous-trading demand premise is behaviourally real rather than hypothetical. | supports | moderate | 0.55 | 0.75 | 0.41 | 2.00 | +0.286 | Observed fact | Daily brief 2026-09-29 (claim on SEC-approved tokenized stock trading) | 2026-09-29 added 2026-10-01 |
| 16 | Retail search attention to stablecoins has surged into the top theme movers (Digital Payments & Fintech +65.7) and is dominated by definitional/first-contact queries, consistent with the leading edge of mainstream awareness in the payments leg. | supports | weak | 0.45 | 0.60 | 0.27 | 1.30 | +0.071 | Observed fact | Attention movers, 2026-10-01 | 2026-10-01 added 2026-10-01 |
| 17 | Policy rates and term premium rose further (10y 5.33%, 30y >5.6%, 5y >5%), deepening the stablecoin/tokenised-MMF reserve float profit pool and the incentive for brokers and issuers to move cash balances on-chain. | supports | moderate | 0.80 | 0.50 | 0.40 | 2.00 | +0.277 | Observed fact | Briefs 2026-09-29 to 2026-10-01 (rates data) | 2026-10-01 added 2026-10-01 |
| 18 | The dollar hit a 16-month high alongside record US tech concentration (Technology rotation +57 against every other sector between -72 and -87), meaning global marginal capital is currently concentrating inside the US legal/market stack rather than diversifying away from it — the opposite of the de-Americanisation mechanism. | contradicts | strong | 0.80 | 0.75 | 0.60 | 0.33 | -0.659 | Observed fact | Briefs 2026-09-29/10-01 (FX and rotation data) | 2026-10-01 added 2026-10-01 |
| 19 | A global term-premium/supply shock occurring despite a soft core PCE print (3.0%) plus an emerging AI credit cycle (Nvidia vendor-financing guarantees, neocloud leverage) raises the odds of a funding/credit stress event in 2027-29 that would stress-test stablecoin reserves and tokenised custody before the trust prerequisite is established. | contradicts | moderate | 0.60 | 0.70 | 0.42 | 0.50 | -0.291 | AI inference | Briefs 2026-09-30/10-01 | 2026-10-01 added 2026-10-01 |
| 20 | Washington's choice of voluntary AI/agent guardrails over statute, following the Clarity Act's failure, confirms a pattern of time-limited and discretionary US regulatory accommodation rather than durable statutory frameworks — the SEC tokenisation exemption is more reversible than a codified regime would be. | contradicts | weak | 0.55 | 0.50 | 0.28 | 0.77 | -0.072 | Observed fact | Brief 2026-09-30 | 2026-09-30 added 2026-10-01 |
Stage of adoption
Unchanged in substance: Eurodollar 1966-70 / ECN 1998-2000 on the securities leg, M-Pesa 2010 / UPI 2018 on the payments leg. The one genuine advance this month is that the securities leg produced its first behavioural print (off-hours share) rather than only a legal carve-out — ECNs circa late 1999 rather than early 1998. Offsetting that, the macro backdrop has moved against the de-Americanisation premise: a 16-month-high dollar and record US equity concentration are the 1990s-style 'capital comes home' regime, not the 2022-style flight regime.
- ✓SEC five-year innovation exemption live, with first observed continuous/off-hours tokenised equity trading activity
- ✓MiCA, Hong Kong/Singapore/UAE/UK digital-securities regimes operative
- ✓Institutional custody and reserve attestation stack in place
- ✓Stablecoin money network effects entrenched; retail awareness now entering mainstream search behaviour
- ✓Mobile-first non-US distribution at scale (Zerodha/Groww, Nubank, Revolut, Binance/Coinbase international)
- ✓Instant domestic rails at majority adoption in the largest account-growth markets (UPI, Pix)
- ✓Reserve-float economics strengthened further by 5.3%+ long-end yields
- ✓Several non-US listing venues individually clearing ~1%+ of global IPO proceeds
- ○Named market makers and published depth in tokenised equities (off-hours share is high but absolute volume remains de minimis)
- ○Atomic cross-venue DvP with enforceable finality across jurisdictions
- ○Prudential/collateral eligibility: no CCP or prime broker yet accepting tokenised money funds as margin
- ○A survived stress event absorbed without user loss
- ○Cross-border interoperability of instant rails at commercial scale (Nexus/mBridge still pilots)
- ○Standardised public reporting to make conditions (b) and (c) verifiable at all — this remains the most under-weighted failure mode
- ○Resolution of capital-control conflicts in India, China, Nigeria
- ○Durable US statute replacing the time-limited exemption
Scenarios
All three conditions clear comfortably by 2029-30. The SEC exemption is imitated by the FCA, MAS, SFC and DIFC, producing mutual-recognition-style tokenised listing venues; a major broker (IBKR, Robinhood International, Coinbase) launches 24/7 tokenised US-equity trading for non-US persons with real market making, and incumbents respond by extending to 24/5 then 24/7, which legitimises rather than kills the format. Stablecoin B2B/remittance value compounds >100%/yr off a low base as Nexus and MiCA-regulated euro/dirham/SGD coins make multi-currency default; tokenised money-market funds become CCP-eligible collateral, pulling institutions across. Non-US venues (HKEX, NSE, Tadawul, Tokyo, plus a Gulf tokenised exchange) each hold >1% of global IPO proceeds and the combined non-US share of new retail accounts is uncontroversially 70%+. Winners: exchanges and brokers with global permissioning, stablecoin-adjacent payment processors, digital custody, key-management security. AI agents become the front end that arbitrages venues on the user's behalf, destroying single-venue pricing power.
Conditions (a) and (c) are met; (b) is met on the listing leg but contested on measurement. Stablecoin settled payment value plus tokenised treasuries/MMF settlement clears $1trn on Artemis/rwa.xyz adjusted methodology by 2029-30, driven overwhelmingly by cash-like instruments and cross-border B2B, not by tokenised equities, which remain a $10-100bn niche used for collateral mobility and off-hours access rather than price formation. Three-plus non-US venues hold measurable listing share — as they already do — but it is continuity of the post-2020 IPO-venue shift rather than a new phenomenon. Non-US platforms open the clear majority of new retail accounts, as they already do, but the US stack retains the assets: dollar-denominated, US-issued securities remain the product even when the wrapper and the app are foreign. The winners are the intermediation layer and the rails, not the challengers to the US asset base; incumbent exchanges and custodians co-opt tokenisation as a back-office upgrade and keep pricing power over data and listings.
A trust shock resets the clock. A top-five stablecoin or a tokenised-custody provider fails a redemption test, or a major offshore venue is hacked at scale, triggering emergency rulemaking, Basel-style punitive capital treatment and retail loss headlines; the SEC exemption is narrowed or allowed to lapse at renewal, and extraterritorial enforcement closes tokenised US-equity access to non-US retail. Simultaneously, 5%+ risk-free rates and a strong dollar keep capital flowing into the existing US stack — record Nasdaq highs and deep liquidity are their own network effect — so no issuer of consequence chooses a tokenised venue. On-chain settlement excluding crypto-native churn stalls in the low hundreds of billions; tokenised equities remain a demo. Non-US retail account growth continues but is domestically ring-fenced by capital controls, so the 'multi-venue, multi-currency' part never materialises for the median user. Incumbent banks, DTCC, NYSE/Nasdaq and card networks keep cross-border pricing power.
Indicators
- ·SEC five-year innovation exemption for tokenised stock trading granted (Sep 2026)
- ·Rising Cloud/AI-agent attention indicating consumer willingness to delegate financial complexity to an interface
- ·~5% policy rates making stablecoin reserve float a large, near-costless revenue pool
- ·Non-US exchanges already clearing 1%+ of global IPO proceeds individually (HKEX, India, Gulf)
- ·UPI/Pix majority domestic adoption and Nexus/mBridge cross-border linkage pilots
- ·Incumbent defensive moves: DTCC/Nasdaq tokenised collateral pilots, extended trading hours
- ·Growth of tokenised money-market funds (BUIDL/BENJI class) as institutional on-ramp
- ·Tokenised equity volumes still de minimis with no real secondary liquidity
- ·Clarity Act stalled — no durable US statutory framework, only a time-limited exemption
- ·Capital controls in India, China, Nigeria blocking the cross-border leg
- ·Basel crypto capital charges and lack of CCP collateral eligibility
- ·Record US equity index highs and a firm dollar pulling capital into the incumbent stack
- ·Continued exchange hacks keeping the trust prerequisite unmet
- ·Bank lobbying and GENIUS-style prohibitions on interest-bearing stablecoins
- ·Whether a major stablecoin or tokenised-custody provider survives its first genuine redemption/stress event
- ·Whether the SEC exemption is renewed/codified or narrowed at review, and whether extraterritorial enforcement blocks non-US offering of US-equity tokens
- ·Interest-rate path: 5% rates make float lucrative but also make bank deposits defend harder and raise the cost of a run
- ·Prudential recognition of tokenised assets as eligible collateral (Basel, CCPs, insurance regulators)
- ·India/China capital-control policy toward outbound retail investing
- ·Whether AI agents become the default financial interface and therefore the venue-selection layer
- ·Speed of incumbent co-option: if NYSE/Nasdaq/DTCC ship 24/7 tokenised settlement first, challengers never get a liquidity foothold
Value pools
| Layer | Emerging need | Opportunity | Becomes | Tickers |
|---|---|---|---|---|
Global multi-venue brokerage / permissioning | Retail and institutions want one account that reaches every venue, currency and hour without a US bank relationship | Brokers with global licences, multi-currency ledgers and 24/5-24/7 books capture order flow and weekend spread; white-label infrastructure to non-US neobrokers becomes a high-margin annuity | gains pricing power | |
Stablecoin issuance, reserve float and on/off-ramps | Dollar access and instant settlement for savers and SME treasurers outside the correspondent-banking chain | Float income at 4-5%, merchant acceptance, and payout rails; incumbent processors defend by embedding stablecoin settlement rather than fighting it | mandatory infrastructure | |
Custody, key management and settlement security | Atomic settlement removes reversibility, so a single key compromise is an irreversible loss — trust must be engineered before convenience is sold | Institutional digital custody, MPC/HSM key management, transaction-screening and travel-rule tooling; identity and fraud defence for 24/7 rails | becomes scarce | |
Exchange listing franchises and market data | Issuers want a global retail bid without ADR friction; investors want reference prices across 24/7 fragmented venues | Consolidated tape, reference pricing, indices and ratings become more valuable as venues fragment — the data layer wins even if any single venue loses | gains pricing power | |
Tokenised fund wrappers and collateral mobility | Institutions want intraday collateral release and yield on idle cash without leaving the regulated perimeter | Tokenised money-market funds and ETFs as the Trojan horse for institutional on-chain settlement; asset managers with distribution capture the wrapper fee | mandatory infrastructure | |
Cross-border FX and correspondent banking margin | SMEs and remitters want 2-6% correspondent cost and T+2 FX to go to near-zero and instant | The margin pool being competed away is the incumbents' highest-return revenue line; card-network cross-border FX spread is the single most exposed profit pool in payments | loses pricing power | |
Bank deposit and broker sweep funding | Savers want cash to earn the policy rate continuously and move instantly | Yield-bearing cash substitutes drain low-cost deposits and sweep spread; funding costs rise for deposit-funded lenders — a new structural NIM risk rather than an opportunity | new risk | |
Agentic interfaces to finance | Multi-venue, multi-currency, 24/7 complexity is unmanageable by humans; the agent becomes the venue-selection layer | Whoever owns the financial agent owns routing and therefore the economics; incumbent distribution plus workflow data beats standalone models | becomes scarce | |
Compliance, surveillance and regulatory reporting for 24/7 markets | Continuous markets have no end-of-day reconciliation window; supervision must become real-time and cross-jurisdictional | Always-on surveillance, sanctions screening and reporting is non-discretionary spend for every venue and broker that goes 24/7 | mandatory infrastructure | |
Always-on settlement compute and low-latency networking | 24/7 global matching and settlement across venues raises baseline compute, storage and network duty cycles | Datacenter capacity, edge points of presence and optical transport for continuously operating financial infrastructure | becomes abundant |
The consensus trade is 'tokenisation disrupts Wall Street'. The more likely outcome is that tokenisation is absorbed as a back-office upgrade by the incumbent US stack while the customer relationship migrates offshore — the asset stays American, the account does not. That inverts who gets hurt. The loser is not NYSE or DTCC, whose listing and clearing franchises are protected by relisting costs and prudential inertia; the loser is the deposit-funded, spread-earning intermediary layer: cross-border FX margin at the card networks, correspondent-banking fees, and above all broker cash sweep, which is a high-single-digit-billion profit pool that exists only because retail cash cannot earn the policy rate continuously. At 5% risk-free rates the incentive to arbitrage that away is maximal, and a programmable dollar that pays its holder is the instrument that does it. Second, if settlement becomes atomic and irreversible, the scarce good flips from execution speed to reversibility and insurance: key management, custody bankruptcy-remoteness, and fraud recovery become the priced services, which is the same 'security replaces convenience' pattern already visible in energy and cyber. Third, and least priced, fragmentation across 24/7 venues makes the reference-price and consolidated-data layer structurally more valuable, not less — the same way Reg NMS fragmentation enriched the data vendors more than the ECNs. Finally, the AI-agent layer is the real venue-selection power: once an agent routes a user's cash, FX and orders across rails automatically, every downstream venue and issuer becomes a commoditised price taker and the agent owner extracts the rent. That is the channel by which Microsoft, Alphabet and workflow incumbents end up with more financial-services economics than any fintech challenger.
- ·Deposit- and sweep-funded brokers and banks whose net interest income depends on retail cash not earning the policy rate (SCHW most exposed, then WFC, BAC, C)
- ·Card networks' cross-border FX spread — the highest-margin line in payments and the exact friction stablecoin rails are priced to eliminate (V, MA)
- ·Correspondent-banking and treasury-services franchises built on nostro/vostro chains (C, JPM at the margin)
- ·Legacy remittance and money-transfer operators outside the universe, and the sponsor banks that serve them
- ·US exchange listing franchises' incremental growth — new listings increasingly go to HKEX, India, the Gulf; existing listings are sticky but the pipeline is not (ICE, CME less exposed than the pure listing franchises)
- ·PayPal-era closed-loop wallets whose moat was cross-border convenience now being commoditised by open rails (PYPL)
- ·Single-jurisdiction, single-currency neobrokers without global permissioning, who face agent-driven price transparency without the venue reach
- ·Any intermediary whose revenue is the calendar — weekend/overnight gapping, market-on-close monopolies, and batch reconciliation services
- ·Compliance-light offshore venues in the bear case: a trust shock concentrates survivors and wipes the long tail
Probability history
Updates log
| Date | Before | After | Δ | New | Retired | What changed | Should change mind? | |
|---|---|---|---|---|---|---|---|---|
| 2026-10-01 | 95.0% | → | 80.3% | -14.7 | 6 | 3 |
| Net: marginally down, and mostly for calibration rather than news reasons. The genuinely new evidence is close to a wash — the off-hours tokenised-volume print and the stablecoin attention surge are real positives for leg (a); the 16-month-high dollar, record US market concentration and the voluntary-guardrails precedent are real negatives for legs (b) and (c), and they directly invalidate the 'jurisdictional diversification' evidence item, which I am retiring. The honest problem is that 95% is too confident for a three-way conjunction in which one leg (>50% of new global retail accounts on non-US platforms in calendar 2031) depends not on behaviour — which is probably already true — but on the existence of a published, comparable global count that does not currently exist and has no owner building it. Add a reversible US exemption, zero prudential collateral recognition, unresolved capital controls, and a macro regime that raises the odds of a trust-destroying stress event inside the horizon, and a posterior in the 88-91% band is better calibrated than 95%. I would mark the probability down ~4-6 points, with the explicit note that the downgrade is driven by verifiability and conjunction risk, not by any deterioration in the underlying adoption mechanism, which continues to track. |