Crypto & digital assets

Crypto & digital assets

Industry ID: 21 | Slug: crypto-digital-assets | Researched: 2026-09-15 | Analyst: agent

Research conditions. The shared WebSearch budget was exhausted before this sector began (one search attempted, returned the budget message). All discovery was performed by direct WebFetch against primary-source URLs. The Tier-A verification below is sound; discovery is narrower than it would be with search, and this sector should be prioritised for re-run alongside 13–18. See §13.

1. Definition and boundaries

In scope. Public blockchain networks (L1 and L2) and the economics of running them; stablecoins as traded and settled tokens; tokenisation of real-world assets and the on-chain money-market complex; decentralised finance including lending, liquid staking, spot DEXs and perpetual DEXs; centralised exchanges, brokers and digital-asset custodians; bitcoin mining and the conversion of mining estates to AI/HPC hosting; central bank digital currencies; exchange-traded crypto products; corporate digital-asset treasury vehicles; and the regulation and enforcement that governs all of it.

Explicitly out of scope, and who owns it.

• Stablecoins inside regulated banking, and the GENIUS Act from the bank side — sector 07. Sector 07 owns tokenised deposits (JPMorgan Kinexys, the 17-bank Clearing House network), the bank deposit-disintermediation debate, and the prudential rulemaking as it lands on bank balance sheets. This dossier covers the token market: supply, issuer economics, reserve quality, circulation and the non-bank charter question. Read §§8–10 of sector 07 alongside this.
• The AI demand driving mining conversion — sector 01. Anthropic and Fluidstack appear here only as counterparties.
• Grid interconnection, generation and power economics — sector 05.
• Payment card networks and consumer fintech — sector 07.
• Semiconductor supply for ASICs and GPUs — sector 03.

Boundary disputes worth naming.

1. Is a stablecoin a payment instrument or trading collateral? The GENIUS Act regulates it as the former. The measured usage (§7, T-21-19) looks substantially like the latter. This is not a definitional quibble: it determines whether this sector belongs next to payments or next to securities market structure.
2. Are converted miners still in this sector? Cipher Mining renamed itself Cipher Digital and now reports in megawatts. At some point these companies leave sector 21 for sectors 01/05. We track them here because the assets, the shareholders and the optionality are still crypto's.
3. Tokenised money-market funds sit on the line between this sector and asset management. The Federal Reserve notes they are already inside existing non-M1 M2 retail money-fund data, i.e. the official statistical system has already decided they are money funds.
4. Perpetual futures. The CFTC now treats certain crypto perpetuals as foreign futures under Regulation 30.1. That is a derivatives-market fact with crypto-market consequences.

2. Subcategories

3. Market structure

Concentration: winner-take-most at every layer, and the winners are not the same firms.

• Stablecoins are a duopoly. USDT and USDC together are roughly 84% of a ~$305bn market (DefiLlama, 2026-09-15). This is more concentrated than most payment networks.
• Crypto ETFs are near-monopolistic. IBIT's $60.62bn (BlackRock, 2026-09-11) is about 64% of US spot bitcoin ETF assets. The distribution advantage of the largest asset manager transferred intact into a new asset class in under three years.
• Tokenised Treasuries are oligopolistic. Five products are roughly 65% of $15.75bn.
• DeFi is concentrated by protocol and by chain. Lido ($24.3bn) and Aave ($18.2bn) are nearly half of $88.2bn TVL; Ethereum is 57% of chain TVL.
• Perp DEXs are one-firm-dominant. Hyperliquid is ~36% of tracked perpetual DEX volume.
• Mining is fragmented but consolidating around power access, not hashrate.

Where margin actually sits. Not where the narrative says. It sits in reserve income on float and in distribution, and the split between them is disclosed: Circle earned $1,320.2m of reserve income in H1 2026 and paid $815.8m — about 58 cents of every revenue dollar — to distribution partners, keeping $103.5m of net income (10-Q, 2026-08-05). The economics of the dollar token accrue to whoever owns the customer, not to whoever issues the token. Coinbase's stablecoin revenue of $292.1m in Q2 2026 is the other side of that ledger.

Margin does not sit in trading: Coinbase's transaction revenue fell to $599.2m in Q2 2026 and the company posted a net loss of $359.5m against net income of $1,428.9m a year earlier.

Barriers to entry. Rising and now largely regulatory. A US national trust bank charter, a MiCA authorisation, or an ETF listing with a qualified custodian are each multi-year, multi- million-dollar undertakings. The 2017–2021 barrier — engineering — is now near zero. The 2026 barrier is a licence and a distribution agreement.

Pricing power. Held by distributors (exchanges, brokers, wallets) over issuers; by the largest ETF sponsor over every other sponsor; and by AI tenants over converted miners, whose alternative buyer is a commodity at a 45% twelve-month drawdown.

Market-size figures, with modeller and date named. We decline to give a single "market size" for this sector because the available figures are definitionally incompatible. The defensible ones are: stablecoin market capitalisation ~$305bn (DefiLlama, 2026-09-15), ~$317bn (Federal Reserve staff, 2026-04-06), ~$320bn (BIS, end-May 2026), $316.6bn (Artemis, 2026-09-15), $304.58bn (RWA.xyz, 2026-09-14); tokenised US Treasuries $15.75bn across 101 products (RWA.xyz, 2026-09-15); DeFi TVL $88.166bn (DefiLlama, 2026-09-15); US spot bitcoin ETF assets ~$95bn (issuer pages plus ETF Database, 2026-09-11/14). Any "$X trillion by 2030" figure encountered in this sector should be treated as forecast and attributed — including Treasury Secretary Bessent's view, cited in Federal Reserve staff research, that the stablecoin market "could grow tenfold by the end of the decade."

4. Who matters

Leading companies. Coinbase Global (COIN, coinbase.com) · Circle Internet Group (CRCL, circle.com) · Tether Holdings (tether.to) · BlackRock (blackrock.com) · Fidelity Investments · Grayscale Investments (grayscale.com) · Strategy Inc (MSTR, strategy.com) · Franklin Templeton · WisdomTree · TeraWulf (WULF, terawulf.com) · Cipher Digital (CIFR, cipherdigital.com) · Core Scientific (CORZ, corescientific.com) · IREN · Galaxy Digital.

Notable startups and protocols. Securitize (securitize.io) · Ondo Finance (ondo.finance) · Paxos (paxos.com) · Ethena Labs (ethena.fi) · Hyperliquid (hyperliquid.xyz) · Lido (lido.fi) · Aave (aave.com) · Sky, formerly MakerDAO (sky.money) · Phantom Technologies · Fireblocks · Chainalysis · Aster · Lighter.

Active investors. a16z crypto (a16zcrypto.com) · Paradigm (paradigm.xyz) · Pantera · Polychain · Haun Ventures · Dragonfly. Caveat: no 2026 fund close or round in this sector was verifiable from primary sources in this research window, and the sector-11 correction applies — record deployment across venture has not produced cash (2021-vintage DPI 0.05x).

Platforms and standards bodies. Ethereum Foundation · Solana Foundation · The Clearing House (tokenised deposit network, sector 07) · ISO 20022 · BIS Innovation Hub Project Agorá.

Regulators. SEC (sec.gov) · CFTC (cftc.gov) · Federal Reserve Board (federalreserve.gov) · OCC (occ.gov) · FDIC · FinCEN (fincen.gov) · NCUA · ESMA (esma.europa.eu) · EBA · ECB (ecb.europa.eu) · FCA (fca.org.uk) · MAS (mas.gov.sg) · HKMA (hkma.gov.hk) · SFC Hong Kong · FSA Japan (fsa.go.jp).

Research institutions. Bank for International Settlements (bis.org) — the 2026 Annual Economic Report chapter is the best single analytical document on this sector · Federal Reserve Board FEDS Notes — six 2026 notes on stablecoins and digital money · Financial Stability Board.

Trade organisations. Blockchain Association (theblockchainassociation.org) · The Digital Chamber (digitalchamber.org) · Global Digital Finance · Crypto Council for Innovation. All are advocacy bodies; their figures are marketing or estimate by default, never fact.

Consumer and civil-society groups. Americans for Financial Reform · Better Markets · Consumer Federation of America · Transparency International (on illicit finance). These are the counterweight to the trade bodies and are systematically under-cited in sector coverage.

5. Products, business models, technologies, customers

Major products. Dollar tokens (USDT, USDC, USDS, PYUSD, USDG); synthetic dollars (USDe); tokenised government money funds (BUIDL, USYC, USDY, iBENJI, WTGXX); exchange-traded products (IBIT, ETHA, FBTC, GBTC); custody and prime brokerage; spot and derivatives trading venues, both custodial and on-chain; liquid staking (stETH); AI/HPC colocation leases sold by former miners.

How money is actually made today.

1. Interest on other people's float. The dominant model. Circle's reserve income was 95% of its H1 2026 revenue. This is a rate-sensitive annuity, which in a sticky-inflation, hawkish-lean environment is a feature — and would invert quickly on cuts.
2. Distribution rent. The fastest-growing and least-discussed model: capturing a share of someone else's float income for owning the customer relationship. 58% of Circle's revenue.
3. Transaction fees. Falling. Coinbase's transaction revenue declined ~21.6% year on year.
4. Asset-management fees. 0.25% on ~$70bn of iShares crypto assets, a conventional and durable business that happens to hold an unconventional asset.
5. Contracted power and space. The mining industry's new model: 20-year leases instead of a volatile commodity.
6. Protocol fees passed through to depositors. Ethena's annualised fees of $284.08m against net protocol revenue of $3.91m — nearly all of it is passed to users, which is the point and also the fragility.

How that is changing. From volatility-linked to rate-linked, and from fee-per-trade to rent-per-balance. The most important structural shift in the sector is that its two flagship public companies are now levered to the federal funds rate rather than to bitcoin's price.

Technologies that matter. Public ledger settlement finality; ERC-20 and equivalent token standards; on-chain order books with off-chain matching (Hyperliquid's architecture); delta-neutral collateral engineering (Ethena); qualified custody with MPC or HSM key management; atomic cross-currency settlement (Project Agorá); iXBRL and machine-readable regulatory disclosure (MiCA white papers). Notably absent from the list of things that matter: consensus-algorithm innovation, which has not been a competitive variable for several years.

Customer segments and what they buy on.

• Retail speculators buy on price momentum and access. The largest segment by count, the smallest by balance.
• Retail savers in high-inflation economies buy on dollar access. The most defensible use case and the least-measured; the Federal Reserve recorded retail-sized wallets (≤$1,000) increasing substantially during 2025.
• Crypto-native trading firms buy on collateral efficiency, uptime and fee schedule. The dominant source of on-chain volume.
• Traditional asset allocators buy on wrapper legitimacy: a ticker, a custodian, a daily NAV. This is why the ETF complex works and direct custody does not scale to them.
• Corporate treasuries — a segment created by Strategy and now in retreat (§10).
• AI laboratories buy megawatts on speed-to-energisation. A customer segment this sector did not have three years ago and which now sets the marginal price of its largest physical asset.

6. Geography

Regulation concentrates in three blocs with three different theories. The US has chosen issuer licensing plus securities-law exemption: the GENIUS Act regime for stablecoins, and an SEC token taxonomy (Release 2026-30) plus proposed Regulation Crypto Assets (2026-08-18) for everything else. The EU has chosen comprehensive activity licensing: MiCA's 18-month grandfathering ended 2026-07-01, after which authorisation is mandatory, with five public registers updated weekly. The UK has chosen a slower, more restrictive path: final FCA rules published 2026-06-30 but permissions only from 2027-10-25, with the retail derivatives and ETN prohibition maintained on the stated basis that retail consumers cannot reliably value the risk.

Capital concentrates in the US, and increasingly in US public markets rather than venture: the $810m Cipher Digital secured notes and Strategy's $17.06bn of ATM issuance are larger than most crypto venture rounds. Sector 11's correction applies with force here — aggregate dollars are carried by a handful of vehicles.

Demand is genuinely global and dollar-denominated. This is the sector's defining geographic fact: 99.4% of fiat-backed stablecoin valuation is US-dollar-pegged (BIS, end-May 2026), and all non-USD stablecoins together total roughly $700m against ~$305bn. Every jurisdiction is building national infrastructure for a market denominated in someone else's currency. The BIS names the consequence directly: "stablecoin dollarisation" in emerging markets as a material financial-stability challenge.

Non-US market covered with a regional source — Japan. Japan's FSA has operated a statutory crypto exchange regime since 2017, longer than any Western jurisdiction, and its experience includes two major exchange failures and recovery. On 2026-02-16 it published the report of the Financial System Council's Working Group on Crypto-asset Systems, the basis for reworking how crypto assets are classified under Japanese financial instruments law, and on 2026-07-23 published guidance on cybersecurity in crypto-asset businesses (fsa.go.jp/en/news/). Japan is the control case for the argument that regulation kills the market: it has had the strictest regime for longest and still has a functioning one. Caveat: only the FSA's English index was retrievable; individual English detail pages 404'd, and the Japanese originals were not machine-translated in this window.

Physical assets concentrate in the US, specifically in Texas and upstate New York — Cipher Digital's 900 MW San Antonio site and 2.5 GW of planned bring-your-own-generation, TeraWulf's Lake Mariner 750 MW campus. This is the only part of the sector with a real geography, and it is being pulled into sectors 01 and 05.

Absent from this dossier: Hong Kong (Stablecoins Ordinance licensing, Project Ensemble) and Singapore (Project Guardian, Global Layer One). Both regulators' sites were inaccessible to automated retrieval — HKMA and SFC returned empty JS-rendered bodies, MAS returned HTTP 403 on every path. This is a real Asian coverage gap, recorded in §13.

7. Historical trend patterns

Twenty years of this sector reduce to a repeating structure: a price cycle dressed as a technology cycle, in which each peak produces a permanent institutional residue much smaller than the peak's claims and much larger than the trough's despair.

The cycles.

• 2011–2013, first bubble. Mt. Gox; the "digital gold" thesis appears.
• 2013–2015, first winter. Mt. Gox fails with ~850,000 BTC. Residue: the first regulatory attention.
• 2017, the ICO wave. ~$6bn raised on whitepapers. Resolution: the SEC's DAO Report and a near-total wipeout. False positive: "the token is the business model." It was not, and this is precisely what Regulation Crypto Assets is now trying to re-legalise with guardrails.
• 2020–2021, DeFi summer and the NFT peak. Resolution: NFT volumes collapsed by well over 90% and never recovered. False positive: "ownership of digital objects is a new asset class."
• 2022, the credit collapse. Terra/Luna, Three Arrows, Celsius, FTX. Residue: proof-of-reserve practice, and a permanent scepticism toward rehypothecation that the CFTC's 2026 decision to permit customer-asset transfer to an affiliated foreign board of trade with re-use rights quietly tests again.
• 2023–2025, the institutionalisation wave. Spot ETFs approved January 2024; the GENIUS Act July 2025; a US administration supportive of the sector. Residue, and it is real: ~$95bn of ETF assets, $15.75bn of tokenised Treasuries, two SEC-registered public companies with audited accounts, and a statutory stablecoin regime.
• 2026, the current phase. Regulatory clarity arriving simultaneously with a ~45% twelve-month price drawdown. This is the first cycle in which the policy and the price have moved in opposite directions, and it is the single most diagnostically useful fact in this dossier.

Specific false positives in THIS sector, named.

1. "Enterprise blockchain" (2016–2019). Every major bank ran a consortium. Nothing survived at scale. Note that the 2026 tokenised-deposit consortium (sector 07) is structurally the same bet with better plumbing.
2. "The flippening." Ether overtaking bitcoin has been imminent since 2017. ETHA's twelve-month return of −37.2% against IBIT's −45.6% is the closest it has come, and it came via a bear market.
3. "Web3 will replace the consumer internet" (2021–2022). Measurable user counts never materialised. Sector 17 covers what actually happened to social platforms.
4. "Institutional adoption is coming" as a standing claim, which has been true and imminent for nine consecutive years. The correct response is to ask which instrument, with what AUM, marked by whom.
5. "Tokenisation of everything." Still asserted; still, in 2026, ~97% concentrated in dollars and short-dated government paper.

The one pattern that has NOT repeated: every prior drawdown was accompanied by regulatory hostility, which let the sector attribute the drawdown to policy. In 2026 the policy went the sector's way and the price fell anyway. That removes the sector's standard explanation and is the strongest available evidence that the price cycle was never primarily about regulation.

8. What is changing now (as of 2026-09-15)

1. The price and the policy have decoupled, and the policy won. Bitcoin is down 45.6% over twelve months and 32.97% year to date on the IBIT benchmark, with the CME CF Bitcoin Reference Rate at $77,209.21 on 2026-09-11; ether is down 46.9% year to date. Over the same period the SEC published a token taxonomy, proposed a registration exemption regime and preempted state securities registration; the CFTC permitted listed spot crypto, classified perpetuals as foreign futures and issued a 24/7 trading advisory; and five federal agencies advanced the stablecoin regime. The most favourable regulatory year in the sector's history coincided with one of its worst price years. Every thesis that treated regulation as the binding constraint has been falsified.

2. Stablecoin supply has stopped growing. The Federal Reserve measured $317bn on 2026-04-06 and observed growth "flattened in Q4 2025 and Q1 2026"; the BIS measured ~$320bn at end-May; three trackers put it at $304.6–316.6bn in mid-September. Five months, no net growth. This is the sector's most important unreported number.

3. The corporate bitcoin treasury model has reversed. Strategy reported a Q2 2026 net loss of $8.22bn, holds 843,775 BTC at a cost of $63.69bn against a market value of $54.77bn, raised its STRC preferred dividend to 12.00%, is repurchasing that preferred at a 13% discount to par, and has a BTC Monetization Program that sold $218.4m of bitcoin year to date, primarily to fund preferred dividends. The flywheel is running backwards, disclosed by the company itself.

4. The mining industry is leaving. TeraWulf signed a 20-year, 410 MW lease with Anthropic. Cipher Mining became Cipher Digital, raised $810m of secured notes, and now reports in megawatts. Core Scientific remained independent after its acquisition did not complete. The assets and the balance sheets are migrating to sectors 01 and 05, taking bitcoin's long-run security budget with them — a structural change nobody in the sector is discussing.

5. The macro brief bites here in a specific way. The FOMC's sticky 3.50–3.75% and its explicit flag on AI-firm valuations and leveraged infrastructure financing are not background for this sector — they are the two live risks. The rate sustains stablecoin and tokenised-fund revenue and suppresses the risk asset. The AI-financing flag lands directly on the 20-year leases and $810m secured notes that converted miners have just signed at what may be an AI capex peak. And per sector 11, aggregate capital figures conceal a median experience that is far worse.

6. What did NOT happen on schedule, consistent with the macro brief's deadline-slippage warning: the GENIUS Act regime is still not in force. Rulemaking across five agencies remains at proposal stage, and the sector's largest commercial question — whether affiliates may pay yield on stablecoins — is unresolved. Any 2026 vendor claim of "GENIUS compliance" describes a posture toward a proposed rule, not compliance with an operative one.

9. The five lists

Five most important current trends

1. T-21-01 — Stablecoin supply plateaus near $305bn after two years of compounding growth
2. T-21-03 — The US securities perimeter for tokens redrawn by rule instead of by litigation
3. T-21-05 — Bitcoin miners convert to AI data-centre landlords, with AI labs as named tenants
4. T-21-04 — Spot crypto ETFs hold ~$95bn but are in a severe drawdown, testing whether holders stay
5. T-21-02 — The GENIUS Act regime is being built by rulemaking and is not yet in force

Five fastest-growing signals (rate of change, not importance)

1. T-21-05 — Mining-to-AI conversion: 410 MW in one lease, $810m in one notes offering, in one quarter
2. T-21-11 — Synthetic dollars: USDe +16% and Ethena TVL +19.3% in 30 days inside a flat market
3. T-21-03 — SEC rulemaking velocity: a taxonomy in March and a full proposed regulation by August
4. T-21-10 — Perpetual DEXs: $631.6bn of 30-day volume and $23.6bn of open interest
5. T-21-09 — CFTC onshoring: four substantive staff actions in a single year

Five trends most likely to affect businesses

1. T-21-02 — GENIUS rulemaking determines who may issue a dollar token and on what terms
2. T-21-08 — The shift from trading fees to distribution rent changes every crypto P&L
3. T-21-06 — Tokenised Treasuries change corporate and trading-firm collateral management
4. T-21-05 — Converted mining capacity is now a material, under-counted source of AI data-centre megawatts
5. T-21-12 — 24/7 market structure will reach every broker, exchange and back office, not just crypto ones

Five trends most likely to affect consumers

1. T-21-04 — ETFs are how ordinary savers hold this asset, and they have just lost ~45% in a year
2. T-21-19 — The payments promise consumers are being sold rests on a volume metric that needs adjusting
3. T-21-15 — A digital euro, UK permissions and Japanese reclassification determine consumer access outside the US
4. T-21-16 — Retail holders of DAT common and preferred are exposed to a model now funding dividends by asset sales
5. T-21-20 — Manufactured metrics are how retail is persuaded; the Gotbit judgment shows it is prosecuted, not theoretical

Overlaps, stated explicitly. T-21-05 appears on the "most important" and "fastest-growing" lists because its capital velocity and its strategic consequence coincide — unusual, and a reason to take it seriously. T-21-02 and T-21-03 are two halves of one US regulatory reset and should be read together; they are separated because the instrument differs (agency rulemaking under a statute versus SEC interpretation and exemptive rulemaking). T-21-04 appears on both the "most important" and "consumer" lists for the same reason: it is simultaneously the sector's best adoption evidence and most households' actual exposure. T-21-19 and T-21-20 are both measurement-integrity records and overlap substantially; T-21-19 concerns one specific inflated metric, T-21-20 the general class.

10. Overhyped / overlooked / cooling / reversing

Most overhyped

1. Stablecoin transaction volume as evidence of payment adoption (T-21-19). The specific evidence that hype outruns substance: the BIS puts 2025 stablecoin transaction volume at ~$28tn and immediately contextualises it as less than three weeks of settlement on a major US wholesale payment system — while an on-chain tracker reports $7.33tn in a single month, implying ~$88tn annualised, a threefold divergence with no published reconciliation. Visa, a company with every incentive to understand this, publishes an adjusted series that strips out addresses exceeding 1,000 transactions or $10m in 30 days, exchange flows, mint/burn, bridges, bots and MEV — an adjustment that exists because the raw number is not a payments number. And the Federal Reserve has declined to include stablecoins in M1 or M2, stating that "few payment stablecoins are currently in operation, making the assessment of their functional use difficult." Three institutions with no commercial stake all say the headline figure needs deflating before it means anything. Nobody has published by how much.

2. "Institutional adoption" evidenced by announcements, TVL and holder counts (T-21-20). The specific evidence: one tracker reported total RWA holders rising 111.23% in thirty days — a doubling in a month, which is a distribution event, not adoption. The same publisher reports "represented asset value" of $361.91bn alongside "distributed asset value" of $38.89bn for the same market, a ninefold spread between two definitions that commentary routinely collapses. Three independent stablecoin trackers disagree by 4% on the sector's single most-cited number. And the SEC's Gotbit judgment (LR-26598, proposed final judgment filed 2026-07-28, with a parallel criminal guilty plea) establishes that wash trading to "create the false impression of market interest" is a prosecuted practice in this sector, not a theoretical risk. The operative rule: a claim counts only if it appears in an audited filing, a regulator's document, a fund's daily NAV, or posted margin.

Most overlooked

1. Stablecoin supply has stopped growing, and it stopped before the regulatory regime took effect (T-21-01). Five months of Federal Reserve, BIS and three-tracker data show no net expansion between April and September 2026. Attention has missed it because the narrative frame is "GENIUS Act unlocks growth" and the data arrived during the unlocking. This single fact undermines the revenue model of most of the sector's private companies and is not, as far as this research could determine, being discussed anywhere.

2. Bitcoin's long-run security budget is being bid away by AI (T-21-18, T-21-05). Every serious miner is reallocating megawatts to twenty-year AI leases because contracted demand beats a commodity at a 45% drawdown. The bitcoin community discusses the security budget in terms of the 2028 halving and fee markets; the actual near-term mechanism is that the marginal megawatt now has a better-paying customer. Missed because the mining-to-AI story is covered as an equity story.

3. Distribution, not issuance, captures stablecoin economics (T-21-08). Circle pays out 58% of revenue to distribution partners. That number is in a 10-Q, is not disputed, and changes the answer to "who wins in stablecoins" completely — it is whoever owns the customer. Overlooked because it is a cost line rather than a headline.

4. The Fed's proposed "payment account" (T-21-14). A non-interest-bearing, no-credit, no-discount-window settlement account for non-federally-insured institutions would resolve a decade of master-account litigation and change the reserve-holding arithmetic for the whole stablecoin regime. It was published on 2026-05-20 without mentioning stablecoins, trust banks or crypto once, which is exactly why almost nobody in this sector noticed it.

Trends that appear to be cooling

1. The digital asset treasury model (T-21-16). Indicator that turned: Strategy initiated repurchases of its own STRC preferred at a 13% discount to par on 2026-07-27 and created a BTC Monetization Program that sold $218.4m of bitcoin year to date, primarily to fund preferred dividends. A company buying back its own paper at a discount while selling the asset it exists to accumulate has stopped being able to issue above NAV. The preferred coupon at 12.00% is a distress-level cost of capital.

2. DeFi TVL and staking-linked revenue (T-21-17). Indicator that turned: Coinbase's blockchain rewards revenue fell 42% year on year to $83.3m — an audited figure, immune to the price-times-quantity problem that makes TVL unreadable. TVL itself is $88.2bn, but with ether down 37% over twelve months a large part of that series is price, not deposits.

3. Standalone bitcoin mining as a business (T-21-18). Indicator that turned: Cipher Mining renamed itself Cipher Digital. A company removing its core activity from its own name, and raising $810m of secured notes against data-centre assets rather than equity against mining cash flow, is the cleanest possible statement of where its operators think the economics are.

Trends that may reverse, and the mechanism

1. The regulatory liberalisation (T-21-03) — the most reversible thing in this dossier. Mechanism: almost all of it is agency interpretation, proposed rules and staff no-action letters, not statute. An interpretation can be withdrawn; a staff letter binds only its recipient and can be rescinded without notice-and-comment; and Regulation Crypto Assets is a proposal whose state-law preemption will be litigated by state securities regulators. The 2026 pattern established across sectors 09, 14 and 16 — courts and states, not federal policy, setting the boundary conditions — applies here with full force. Falsifiable early indicator: whether Regulation Crypto Assets is adopted substantially as proposed, and whether any state AG files against the preemption provision.

2. The mining-to-AI pivot (T-21-05, T-21-18). Mechanism: an AI capex slowdown. These companies are signing twenty-year leases and issuing secured notes against a demand curve the FOMC has explicitly flagged for "high AI-firm equity valuations and increased leveraged financing of infrastructure buildout." If that demand digests, converted miners will have abandoned one business without securing the other — and bitcoin, having recovered ~19% from its July trough, would be the business they abandoned. Falsifiable early indicator: lease renegotiations, or a miner publicly re-energising hashrate on converted capacity.

3. Stablecoin supply (T-21-01). Mechanism, benign: GENIUS taking effect with permissive affiliate-yield treatment would let issuers compete on rate and could restart growth sharply. Mechanism, adverse: a rate cut compresses reserve income at the same moment compliance costs land. Falsifiable early indicator: the affiliate-yield provision in the final rules, and whether the supply series breaks $320bn or $290bn first.

4. The whole sector's price cycle. Mechanism: it has fully reversed four times in fifteen years and the current drawdown is not unusual by historical standard. Anyone treating the 2026 drawdown as structural should note bitcoin recovered from $64,915 (2026-07-27) to $77,209 (2026-09-11) inside six weeks.

11. Risks and major uncertainties

Sector-specific risks.

• Issuer concentration. Two stablecoin issuers are ~84% of supply; one ETF is ~64% of US spot bitcoin ETF assets; one custodian holds most ETF bitcoin. Each is a single point of failure for a systemically relevant share of the sector.
• Reserve quality asymmetry. Federal Reserve staff found USDT holds ~1.04x reserves per coin but only ~0.74x in higher-quality assets, against 1.0x for USDC. The market treats both as "a dollar." A stress event would discover they are not.
• Rehypothecation, again. CFTC Staff Letter 26-17 permits transferring customer-owned digital commodities and payment stablecoins to an affiliated foreign board of trade with re-use rights. That structure is a recognisable ancestor of 2022's failures.
• Synthetic-dollar reflexivity. USDe's yield depends on positive perpetual funding. Funding inverts. A depeg there would be reported as a "stablecoin failure" and contaminate the regulated segment that had nothing to do with it.
• AI counterparty and duration risk in mining. Twenty-year leases to a small number of private AI laboratories, financed with secured debt, at a point the FOMC has flagged.
• DAT contagion. Strategy has a $3.75bn reserve. Its imitators do not.
• Measurement risk. The sector's headline metrics are manufacturable and demonstrably manufactured (LR-26598). An intelligence pipeline that ingests aggregator dashboards uncritically will import fabricated data at scale.
• Regulatory reversal. Most of 2026's liberalisation is sub-statutory.

Genuine unknowns — and the distinction matters.

Things we don't know but could find out: the payment-versus-collateral split of stablecoin volume (Visa's adjusted series exists but did not render figures on fetch; Allium holds the underlying data); the MiCA authorised-CASP count (the register exists as weekly CSVs; the download path could not be resolved); Hong Kong and Singapore's current licensing state (regulator sites blocked); Tether's current attested reserve composition (published quarterly; the live dashboard is JS-rendered); bitcoin network hashrate and hashprice (published; the source fetched returned 404). Every one of these is a retrieval failure, not an epistemic limit, and every one is fixable with a headless browser or a licensing budget.

Things nobody can know: whether stablecoin supply resumes growing or has structurally topped; whether the ETF holder base converts to a redeemer base if the drawdown extends; whether AI compute demand sustains twenty-year leases; whether a future SEC rescinds the token taxonomy; whether the EU adopts digital euro legislation in the remaining months of 2026; and the counterfactual that would settle the sector's central argument — what the 2026 price would have been without the regulatory liberalisation.

12. Scenarios to 2030

Base — "a regulated, rate-levered, smaller-than-promised financial utility." Stablecoin supply oscillates in the $250–450bn range; the GENIUS regime takes effect in 2027 with restrictive affiliate-yield treatment; tokenised Treasuries reach $40–80bn as collateral infrastructure; ETFs persist as the retail and institutional access layer; two or three converted miners become successful data-centre REIT-alikes and the rest are acquired. The sector becomes boring, profitable for a handful of firms, and considerably smaller than its 2025 forecasts. Early indicator: stablecoin supply holding a $280–330bn band through 2027 without breaking out.

Upside — "the payment use case becomes real." Adjusted (not raw) stablecoin payment volume demonstrably grows in cross-border corridors; affiliate yield is permitted and issuers compete on rate, pulling genuine balances from deposits and money funds; supply passes $600bn; tokenised Treasuries pass $100bn as the collateral standard. Early indicator: Visa's adjusted volume series growing materially faster than its raw series — the only metric that would distinguish real payments from more collateral.

Downside — "the drawdown becomes structural." Bitcoin fails to recover; ETF assets fall below $50bn on redemptions rather than marks; Coinbase posts consecutive losses; several DATs fail and are forced sellers; stablecoin supply contracts below $250bn as trading collateral unwinds; the GENIUS regime arrives to regulate a shrinking market. Early indicator: ETF share count falling, not just net assets — that separates redemption from mark-down, and it is published daily by every issuer.

Disruption — "tokenised bank money wins." The 17-bank Clearing House network (sector 07) and JPMorgan's Kinexys deliver institutional settlement inside the banking system with deposit insurance, central bank money at the base and no reserve-quality question. Stablecoins are relegated to retail and offshore. The BIS's "unified ledger" becomes the institutional architecture and Project Agorá its proof of concept. Early indicator: tokenised deposit daily volume (already >$7bn at Kinexys per sector 07) exceeding adjusted institutional stablecoin settlement volume.

Regulatory — "the reversal." A change in Commission composition rescinds the token taxonomy; state AGs defeat the preemption provision; a court vacates Regulation Crypto Assets on the same statutory-authority theory that felled the CFPB's Section 1033 rule and the IEEPA tariffs. Issuance returns offshore. Early indicator: the first state-AG complaint against the preemption provision, or a Commission vote on Regulation Crypto Assets falling on party lines.

Failure — "a solvency event at the centre." A reserve-quality event at the largest issuer, a funding-rate inversion breaking a synthetic dollar, or a rehypothecation failure at a US-affiliated offshore venue under the new Part 30 structure. Given 84% issuer concentration and the documented 0.74x high-quality reserve ratio, this is the tail that matters. Contagion would run through tokenised collateral into the derivatives complex the CFTC has just connected to US registrants. Early indicator: a sustained secondary-market discount on the largest stablecoin — the BIS already notes that "secondary market prices of stablecoins to date deviate from par."

13. Data gaps and limitations

Discovery was severely constrained. The shared WebSearch budget was exhausted before this sector began. One search was attempted and returned the budget message; all subsequent research was direct WebFetch against primary-source URLs. Tier-A verification is sound — this dossier rests on SEC filings, Federal Reserve and BIS research, CFTC and Fed releases, ESMA, ECB, FCA and FSA pages — but discovery is narrow, and secondary corroboration from FT, Reuters and Bloomberg is almost entirely absent. Recommend re-run at full search budget, alongside 13–18.

Could not verify — access blocked or unrenderable:

• MiCA authorised-CASP count. The register exists as five weekly CSVs; the dated /sites/default/files/ path 404'd and no stable download URL could be resolved. T-21-07 is therefore single_source with no authorisation count asserted.
• Hong Kong and Singapore entirely. HKMA and SFC pages returned empty JS-rendered bodies or 404s; MAS returned HTTP 403 on every path attempted. Stablecoin Ordinance licensing, Project Ensemble and Project Guardian are missing from this dossier. This is the largest single geographic gap.
• Visa Onchain Analytics figures. The dashboard returned its adjustment methodology but no numbers (JS-rendered). This is the single most valuable missing dataset in the sector: it is the only credible adjusted payment-volume series, and its absence is why T-21-19 can establish that the headline figure is inflated but not by how much.
• Tether's current attested reserves. The live balance page rendered no figures. The USDT supply discrepancy between trackers (~$10bn) is therefore unreconciled.
• OCC entirely. Every occ.gov and occ.treas.gov path returned ROBOTS_DISALLOWED with robots.txt itself failing or timing out. National trust charter decisions are sourced second-hand via sector 07.
• Federal Register API returned PROXY_REJECTED (HTTP 403). The authoritative record of whether the GENIUS regime is in force could not be queried directly.
• Bitcoin network hashrate and hashprice. Source 404'd; T-21-18 rests on corporate behaviour rather than on mining economics series.
• IREN, Galaxy Digital, Grayscale — IR pages 403'd or returned no figures.
• Developer activity for DeFi was not measured at all.

Conflicting figures, recorded rather than resolved (10 contradictions across the trend records):

1. Total stablecoin market cap: $304.58bn (RWA.xyz) / $305.10bn (DefiLlama) / $316.6bn (Artemis), same week. Token inclusion rules differ (122 vs 155 tokens tracked).
2. USDT supply: $183.37bn (DefiLlama) vs $193.3bn (RWA.xyz) — a ~5% gap, most plausibly authorised-but-not-issued treasury tokens or chain coverage.
3. Annual stablecoin transaction volume: ~$28tn for 2025 (BIS) vs ~$88tn annualised from RWA.xyz's monthly figure. Almost certainly adjusted-versus-raw. This gap is the measurement problem.
4. Tokenised RWA market size: $38.89bn "distributed" vs $361.91bn "represented" — two published definitions from one source, differing ninefold.
5. UK retail access to crypto ETNs: the FCA's standing guidance page (retrieved 2026-09-15) states the retail ETN prohibition is maintained, which conflicts with widely reported 2025 policy change. Only the FCA page was retrievable as primary; unresolved.
6. Plus USDe supply, IBIT net assets, bitcoin price reference, and two perpetual-volume aggregates — all timing or definitional, all recorded in the records.

Stale-data trap encountered and avoided. The most widely cited public bitcoin-ETF flow tracker returned data ending 2025-04-22 — sixteen months stale — while presenting as current. It was discarded, and ETF evidence was rebuilt from issuer product pages. This is Addendum 2's "date laundering" warning materialising exactly as described.

Methodological limitations. (a) On-chain aggregators are Tier B at best; they are not audited and their inclusion rules are undisclosed. Where a claim is load-bearing, this dossier pairs them with a 10-Q, a Federal Reserve note or a BIS figure. (b) Holder counts throughout are addresses, not people. (c) Non-English sources: Japan's FSA is cited from its English index only; the Japanese originals were not consulted. (d) No consumer survey data on crypto ownership was obtained for any jurisdiction.

14. Ranking scorecard

15. Sources

1. "Stablecoins dashboard — total market cap and per-issuer supply", DefiLlama, https://defillama.com/stablecoins, 2026-09-15, B
2. "Stablecoins in 2025: Developments and Financial Stability Implications" (Carapella, Lubis, Vardoulakis), Federal Reserve Board FEDS Notes, https://www.federalreserve.gov/econres/notes/feds-notes/stablecoins-in-2025-developments-and-financial-stability-implications-20260408.html, 2026-04-08, A
3. "Anchoring trust in money: innovation beyond stablecoins", BIS Annual Economic Report 2026 Ch. III, Bank for International Settlements, https://www.bis.org/publ/arpdf/ar2026e3.htm, 2026-06-28, A
4. "Banks in the Age of Stablecoins: Lessons from Their Historical Responses to Financial Innovations" (Hempel, Perez-Sangimino, Wang), Federal Reserve Board FEDS Notes, https://www.federalreserve.gov/econres/notes/feds-notes/banks-in-the-age-of-stablecoins-lessons-from-their-historical-responses-to-financial-innovations-20260501.html, 2026-05-01, A
5. "Payment Stablecoins and Cross Border Payments: Benefits and Implications for Monetary Policy Implementation" (Kim, Ruprecht, Styczynski), Federal Reserve Board FEDS Notes, https://www.federalreserve.gov/econres/notes/feds-notes/payment-stablecoins-and-cross-border-payments-benefits-and-implications-for-monetary-policy-20260330.html, 2026-03-30, A
6. "New Forms of Money and the U.S. Monetary Aggregates" (Payne, Styczynski), Federal Reserve Board FEDS Notes, https://www.federalreserve.gov/econres/notes/feds-notes/new-forms-of-money-and-the-u-s-monetary-aggregates-20260904.html, 2026-09-04, A
7. "Federal Reserve Board requests comment on proposal to require certain payment stablecoin issuers to maintain an effective customer identification program", Federal Reserve Board, https://www.federalreserve.gov/newsevents/pressreleases/bcreg20260618a.htm, 2026-06-18, A
8. "Federal Reserve Board requests public comment on a proposal to establish a 'payment account'", Federal Reserve Board, https://www.federalreserve.gov/newsevents/pressreleases/other20260520a.htm, 2026-05-20, A
9. "Agencies clarify the capital treatment of tokenized securities", Federal Reserve Board / FDIC / OCC, https://www.federalreserve.gov/newsevents/pressreleases/bcreg20260305a.htm, 2026-03-05, A
10. "SEC Clarifies the Application of Federal Securities Laws to Crypto Assets" (Release 2026-30, interpretive file S7-2026-09), U.S. Securities and Exchange Commission, https://www.sec.gov/newsroom/press-releases/2026-30-sec-clarifies-application-federal-securities-laws-crypto-assets, 2026-03-17, A
11. "SEC Proposes New Regulation Crypto Assets" (Release 2026-76, File 33-11434), U.S. Securities and Exchange Commission, https://www.sec.gov/newsroom/press-releases/2026-76-sec-proposes-new-regulation-crypto-assets, 2026-08-18, A
12. "Commission Staff Confirms the Categorization of Certain Crypto Asset Perpetuals as Foreign Futures…" (Release 9241-26, Staff Letter 26-17), U.S. Commodity Futures Trading Commission, https://www.cftc.gov/PressRoom/PressReleases/9241-26, 2026-05-29, A
13. "CFTC Staff Issues Advisory on 24/7 Trading, Clearing, and Settlement" (Release 9239-26), U.S. Commodity Futures Trading Commission, https://www.cftc.gov/PressRoom/PressReleases/9239-26, 2026-05-29, A
14. "CFTC Staff Issues No-Action Position to Self-Custodial Crypto Asset Wallet Software Provider" (Release 9197-26, Staff Letter 26-09), U.S. Commodity Futures Trading Commission, https://www.cftc.gov/PressRoom/PressReleases/9197-26, 2026-03-17, A
15. Circle Internet Group Inc., Form 10-Q for the quarter ended June 30, 2026, SEC EDGAR, https://www.sec.gov/Archives/edgar/data/1876042/000187604226000248/crcl-20260630.htm, 2026-08-05, A
16. Coinbase Global Inc., Form 10-Q for the quarter ended June 30, 2026, SEC EDGAR, https://www.sec.gov/Archives/edgar/data/1679788/000167978826000088/coin-20260630.htm, 2026-07-30, A
17. "Strategy Announces Second Quarter 2026 Financial Results", Strategy Inc, https://www.strategy.com/press/strategy-announces-second-quarter-2026-financial-results_07-30-2026, 2026-07-30, A
18. "Strategy Announces Digital Credit Capital Framework, USD Reserve Policy and STRC Dividend Policy", Strategy Inc, https://www.strategy.com/press/strategy-announces-digital-credit-capital-framework_06-29-2026, 2026-06-29, A
19. "iShares Bitcoin Trust ETF (IBIT)" product page, BlackRock / iShares, https://www.ishares.com/us/products/333011/ishares-bitcoin-trust, 2026-09-11, A
20. "iShares Ethereum Trust ETF (ETHA)" product page, BlackRock / iShares, https://www.ishares.com/us/products/337614/ishares-ethereum-trust, 2026-09-11, A
21. "Tokenized U.S. Treasuries" dashboard, RWA.xyz, https://app.rwa.xyz/treasuries, 2026-09-15, B
22. "Stablecoins" dashboard — supply, holders, transfer volume, active addresses, RWA.xyz, https://app.rwa.xyz/stablecoins, 2026-09-14, B
23. "Stablecoins sector overview", Artemis, https://www.artemis.ai/sectors/stablecoins/overview, 2026-09-15, B
24. DeFi TVL, DEX volume and perpetuals dashboard, DefiLlama, https://defillama.com/, 2026-09-15, B
25. Perpetuals / derivatives DEX rankings, DefiLlama, https://defillama.com/protocols/derivatives, 2026-09-15, B
26. Ethena protocol TVL, fees and revenue, DefiLlama, https://defillama.com/protocol/ethena, 2026-09-15, B
27. "Markets in Crypto-Assets Regulation (MiCA)" — implementation, interim register, technical standards, ESMA, https://www.esma.europa.eu/esmas-activities/digital-finance-and-innovation/markets-crypto-assets-regulation-mica, 2026-09-09, A
28. "Digital euro" project pages, European Central Bank, https://www.ecb.europa.eu/euro/digital_euro/html/index.en.html, 2026-07-31, A
29. "Cryptoassets" — UK regulatory regime and final rules, Financial Conduct Authority, https://www.fca.org.uk/firms/cryptoassets, 2026-06-30, A
30. "Publication of the Report by the Working Group on Crypto-asset Systems of the Financial System Council", Financial Services Agency of Japan, https://www.fsa.go.jp/en/news/index.html, 2026-02-16, A
31. TeraWulf newsroom — Anthropic 20-year 410 MW lease and Lake Mariner expansion, TeraWulf Inc., https://www.terawulf.com/news, 2026-07-08, B
32. Cipher Digital investor relations press release archive — rebrand, $810m senior secured notes, Black Pearl, 900 MW San Antonio site, Fluidstack/Anthropic water commitment, Cipher Digital Inc., https://investors.cipherdigital.com/news-events/press-releases, 2026-09-14, A
33. "SEC v. Gotbit Consulting LLC and Fedor Kedrov" (Litigation Release LR-26598), U.S. Securities and Exchange Commission, https://www.sec.gov/enforcement-litigation/litigation-releases/lr-26598, 2026-08-03, A
34. "SEC Announces Agenda and Panelists for Roundtable on Preparations for 24-Hour Trading" (Release 2026-83), U.S. Securities and Exchange Commission, https://www.sec.gov/newsroom/press-releases/2026-83-sec-announces-agenda-panelists-roundtable-preparations-24-hour-trading, 2026-09-01, A
35. "Bitcoin ETFs" assets-under-management list, ETF Database, https://etfdb.com/themes/bitcoin-etfs/, 2026-09-14, C (used only to corroborate issuer-primary figures; never as sole support)
36. Core Scientific Inc. EDGAR submissions index and Form 8-K of 2026-08-27 (revolving credit facility, event date 2026-08-25), SEC EDGAR, https://data.sec.gov/submissions/CIK0001839341.json, 2026-09-02, A
37. CFTC press release index — digital asset actions 2025–2026 (9112-25, 9115-25, 9130-25, 9145-25, 9146-25, 9152-25, 9192-26, 9197-26, 9198-26, 9200-26, 9239-26, 9241-26, 9288-26), U.S. Commodity Futures Trading Commission, https://www.cftc.gov/PressRoom/PressReleases, 2026-08-21, A
38. Chains ranked by DeFi TVL, DefiLlama, https://defillama.com/chains, 2026-09-15, B
39. Real-world asset tokenization overview — asset classes, holders and values, RWA.xyz, https://app.rwa.xyz/, 2026-09-15, B
40. Sector 07 dossier, Finance & Fintech (GENIUS Act rulemaking and tokenised deposits, bank side), Phase 1 research programme (internal), file:///home/claude/research/dossiers/07-finance-fintech.md, 2026-09-15, B
41. Federal Reserve Board 2026 press release index, Federal Reserve Board, https://www.federalreserve.gov/newsevents/pressreleases/2026-press.htm, 2026-09-15, A
42. Federal Reserve FEDS Notes 2026 index, Federal Reserve Board, https://www.federalreserve.gov/econres/notes/feds-notes/2026-index.htm, 2026-09-15, A
43. Visa Onchain Analytics — stablecoin adjusted-volume methodology, Visa Inc., https://visaonchainanalytics.com/transactions, 2026-09-15, A (methodology retrieved; figures did not render — see §13)
44. Circle transparency page (reserve composition and attestation practice), Circle Internet Group, https://www.circle.com/transparency, 2026-09-10, A
45. BIS Annual Economic Report 2026 (chapter list), Bank for International Settlements, https://www.bis.org/publ/arpdf/ar2026e.htm, 2026-06-28, A

More research in this sector · Explore trend records