The SEC proposed new rules for crypto fundraising, Japan admitted its first new exchange registrant in nearly four years, and Texas started checking whether hundreds of enormous electricity requests are real. Three regulators, one week — and, we would argue, one theme.

This week in 30 seconds

1. Regulation. The SEC proposed a tailored crypto-asset offering framework with two exemptions. It is a proposal, open for comment until 20 October.
2. Market access. Laser Digital Japan completed registration as a Crypto Asset Exchange Service Provider. Launch date and service scope are unannounced.
3. Infrastructure. Texas began verifying the credibility of major electricity-connection requests, covering AI data centres and crypto mining alike.

Market pulse

Market data snapshot as of 05:10 UTC on 26 August 2026. Source: CoinGecko. Editorial coverage runs 18–24 August 2026; this market snapshot is a separate, later observation. Stablecoin prices are shown in EUR, so their movements may reflect changes in the EUR/USD exchange rate as well as market pricing. A historical total-market seven-day change is not available from the data provider, so BTC dominance is shown instead as context, not as a seven-day change. Data is indicative and provided for information only.

The permission economy

For most of the past decade the limiting factor in digital assets was money: whoever could raise it could build. This week made a change in that unusually visible.

Three authorities acted independently — the connection drawn here is our analysis, not a policy programme. But each adjusted a different gate: permission to raise capital, to operate in a regulated market, and to draw electricity.

Gate one: permission to raise capital

On 18 August 2026 the US Securities and Exchange Commission issued a proposed rule titled Regulation Crypto Assets (File Number S7-2026-27; Release Numbers 33-11434 and 34-106150). It is a proposal. It is not law, and it is not a final rule.

The framing repays attention. The proposal addresses “certain investment contracts involving crypto assets” — not crypto assets as such. That distinction has been the quiet centre of US crypto regulation. Under the Howey line of cases, what is regulated as a security is the investment contract: the arrangement, the promises made, the expectation of profit from the efforts of others. The token is merely its object. A token sold under a promotional scheme may sit inside a securities offering; the same token traded later between two strangers may not.

Within that framing, the proposal offers two exemptions from the registration requirements of the Securities Act of 1933. The first would permit offerings of up to $5 million during a four-year period. The second would permit offerings of up to $75 million during each 12-month period. Issuers relying on either would provide principles-based disclosures rather than the full registration package, and would remain subject to the antifraud and antimanipulation provisions of the securities laws. The proposal also includes a conditional safe harbour from the term “investment contract” in the definition of “security”.

In practice the $5 million tier is closer to a seed round than a public offering, and the four-year window limits repeat use; the $75 million tier is materially larger. Principles-based disclosure is a genuine shift in posture: it asks issuers to explain what matters about their arrangement rather than complete a fixed form.

What remains unresolved is most of it. Comments are open until 20 October 2026; the Commission must then consider them and adopt a final rule, possibly amended, or decline. Nothing suggests every token or transaction would qualify, and nothing removes the underlying legal test.

The international relevance is indirect but real. Issuers choose venues, and a defined US path for token fundraising would change the relative attractiveness of MiCA, the UAE, Singapore and Hong Kong — not because those regimes weaken, but because the comparison set changes.

Gate two: permission to operate in a regulated market

On 21 August 2026, Laser Digital Japan Co., Ltd. — the Japanese entity of Nomura’s digital-assets subsidiary Laser Digital — announced that it had completed registration and is authorised to operate as a Crypto Asset Exchange Service Provider under Japan’s Payment Services Act.

Japan’s Financial Services Agency publishes a register of crypto-asset exchange service providers. The edition dated 21 August 2026 lists Laser Digital Japan as entry number 00032 under the Director of the Kanto Local Finance Bureau, with a registration date of 21 August 2026, covering six assets: BTC, ETH, XRP, BCH, LTC and SHIB.

It also settles a claim that circulated widely this week. The entry immediately preceding Laser Digital Japan is Binance Japan, number 00031, registered 14 October 2022, and between those dates the register records no new registrations anywhere in Japan. The description of Laser Digital as the first new registrant since 2022 is therefore supportable — but the register proves an absence of entries, not a policy. No authority has announced a freeze, and applications can be withdrawn or abandoned for ordinary reasons.

Registration is also not a launch. Laser Digital says it will initially serve to enhance liquidity across the domestic market for local virtual-asset service providers, with future services to include trading opportunities for institutional investors — launch date and scope of services to be announced later.

Why should banks and asset managers care about a registration that has produced no service yet? Because regulated market access is durable in a way product launches are not. It is granted to a specific legal entity, subject to conditions and supervision; it is not transferable, it can be lost, and it takes time to obtain. Where a market grants few, holding one is a structural position.

Against Europe’s MiCA framework, Japan’s model differs in architecture rather than intent. MiCA provides an authorisation passportable across the EU single market. Japan’s Payment Services Act registration is national, granted through a Local Finance Bureau, and is being reconsidered: legislation passed in 2026 is moving crypto oversight towards the Financial Instruments and Exchange Act, a stricter, securities-style regime. Firms entering Japan now enter under one framework while another is being assembled.

Gate three: permission to draw electricity

The third gate is physical, and the one most likely to bind first. Texas is the centre of gravity for both AI data centres and bitcoin mining, for the same reasons: land, speed of construction and an unusually accessible grid. That accessibility is now being tested. Following a directive from Governor Greg Abbott on 3 August 2026, ERCOT is verifying whether large electricity-connection requests are credible before more projects advance. On 21 August ERCOT told the Public Utility Commission of Texas that it aims to deliver a comprehensive verification report by 10 December 2026, having sought and received relief from an earlier deadline it had missed.

Roughly 300 large loads sit in the first tranche. Facilities of 25 MW and above not yet operating — data centres and crypto-mining sites alike — must supply community-impact information covering power supply, water use, cooling, public incentives and ownership.

The numbers require discipline. More than 1,800 projects sit in ERCOT’s interconnection queue, representing more than 474 GW of potential demand. That figure has been repeated as though it were a forecast of consumption. It is not: a queue is a list of applications. Utilities say many are unlikely to proceed for want of financing or tenants, and the queue’s size is why ERCOT has been unable to publish a reliable demand forecast.

This is also where crypto mining and AI genuinely diverge. A bitcoin miner can stop within seconds and lose only the revenue of those seconds. Many AI training and inference workloads cannot: they carry uptime commitments, and interrupting them is expensive or contractually impossible. That difference has real value to a grid operator, because a large flexible load absorbs shocks during scarcity.

It is also why miners have been converting or diversifying sites towards AI and high-performance computing. The economics are attractive, but conversion trades away the flexibility that made mining useful to the grid. A site that once curtailed on demand and now runs continuous inference is a different kind of neighbour.

Sustainability watch: what 474 GW does and does not mean

Grid flexibility has genuine economic and environmental value. It does not follow that mining is sustainable.

Emissions from any large load depend on four things: the source of the electricity, the location, the timing and the counterfactual — what that power would otherwise have done. A flexible load drawing from a coal-heavy grid at peak is not made clean by its ability to switch off. A load absorbing curtailed wind that would otherwise have been spilled is a materially different proposition. Both are called “flexible”; only one has a strong environmental argument, and which is which depends on evidence rather than category.

For investors assessing miners and AI-infrastructure projects the practical questions are narrower than the headlines: does the project hold an interconnection position or merely an application; is it financed; does it have tenants; what is its power mix; and can it curtail without breaching a contract?

Standards watch: ISSB moves from adoption to implementation

On 18 August 2026 the IFRS Foundation Trustees announced a five-year operating and financing plan covering both the IASB and the ISSB. More than 45 jurisdictions are now using ISSB Standards, and companies in 18 jurisdictions are expected to be issuing reports by 2027.

The significance is the shift in emphasis. The ISSB’s first phase was about persuading jurisdictions to adopt its standards; the next is about supporting the companies that must apply them — slower, less visible work, and more consequential for what investors actually receive. For companies in those 18 jurisdictions, sustainability reporting is moving from a communications exercise towards an assurance-grade disclosure process.

Learn: why a token’s face value is not real liquidity

In late August an attacker exploited a bridge affecting The Sandbox’s SAND token, minting a very large quantity of unbacked tokens. Reported face value ran into tens of billions of dollars; reported proceeds were approximately $665,000 to $675,000. That gap is worth understanding.

Face value is quantity multiplied by the quoted market price. It assumes, silently, that every unit could be sold at that price. For a large unauthorised mint this assumption is false by construction: the tokens were created outside the supply the market had priced.

Circulating liquidity is the quantity that can actually change hands — far smaller than market capitalisation, and spread across venues.

Market depth and slippage describe what happens when you sell into that liquidity. Order books thin quickly: each successive unit achieves a worse price, and beyond a certain size the price collapses toward whatever a buyer will pay for an asset visibly being dumped.

Realised proceeds is the only number that describes an actual loss. In this case, the attacker extracted a sum roughly five orders of magnitude below the face value of what was minted.

The gap between the reported $665,000 and $675,000 is itself instructive: different observers measured at different moments, in a moving market. Neither figure is wrong; both are snapshots.

Two further points. Cross-chain bridges concentrate risk because a message on one chain instructs a mint on another; if that message is forged or the permissions hijacked, supply can be created without collateral. And exchanges may suspend transfers for an affected asset while they establish whether arriving tokens are legitimate — a protective measure, not a judgement on the asset.

Upbit did exactly that. In a notice registered at 11:12 on 22 August 2026, the exchange suspended deposits and withdrawals of SAND on the Ethereum network, citing indications of a security incident and the user-protection provisions of Korea’s Virtual Asset User Protection Act. Trading was not suspended, and the notice states that transfers will resume once stability is confirmed. We report only what that notice says.

What to watch

1. 20 October 2026. Comments close on the SEC’s Regulation Crypto Assets proposal. The informative signal is whether industry responses concentrate on the size of the two offering caps or on the conditional safe harbour, which is the more structurally significant element.
2. 10 December 2026. ERCOT aims to deliver its verification report to the Public Utility Commission of Texas. Watch how many of the roughly 300 large loads survive scrutiny, and whether ERCOT is then able to publish a demand forecast.
3. Date unannounced. Laser Digital Japan’s institutional service scope and launch date. Watch whether the offering extends beyond liquidity provision to domestic operators.
4. Unconfirmed. India’s reported first tokenised corporate bond, said to settle using the Reserve Bank of India’s wholesale central bank digital currency. There the cash leg moves as central bank money on a permissioned ledger and the security leg as a token, allowing delivery-versus-payment: both legs settle simultaneously, so neither party carries the risk that the other fails to deliver. We treat this as a reported plan requiring first-party confirmation from REC, SEBI or the RBI before we would report it as completed.

The takeaway

The industry spent years asking whether the technology works. It largely does. The consequential question now is who is allowed to use it at scale — answered by institutions predating crypto entirely: securities regulators, financial supervisors and the people who run electricity grids.

There is an under-discussed consequence for sustainability. The disclosure grid operators now demand — power mix, water, siting, curtailment capability — is close to what ISSB-aligned reporting is designed to produce. Firms that can already answer those questions credibly will move through the gates faster: not because they are greener, but because they are legible.

Sources

1. US Securities and Exchange Commission — Proposed Rule, Regulation Crypto Assets, File S7-2026-27, Releases 33-11434 and 34-106150, issued 18 August 2026; comments due 20 October 2026. sec.gov
2. Nomura Holdings — “Laser Digital Japan Receives Regulatory Approval To Provide Institutional Crypto Asset Services In Japan”, 21 August 2026. nomuraholdings.com
3. Japan Financial Services Agency — register of Crypto Asset Exchange Service Providers, edition dated 21 August 2026 (entry 00032, Kanto Local Finance Bureau). fsa.go.jp
4. Utility Dive — “ERCOT aims to complete Texas governor’s data center audit by December”, 21 August 2026. utilitydive.com
5. The Texas Tribune — “Texas will audit up to 300 projects, mostly data center proposals”, 14 August 2026 (background, outside the reporting window). texastribune.org
6. IFRS Foundation — “Trustees announce five-year plan”, 18 August 2026. ifrs.org
7. Upbit — notice “샌드박스(SAND) 입출금 일시 중단 안내”, registered 22 August 2026 11:12. upbit.com
8. Market data — CoinGecko, retrieved 05:10 UTC on 26 August 2026.

Company and institution names identify the publishers of the sources cited. Mugano is not affiliated with, and does not endorse, any third-party product mentioned.

More from Mugano

The Mugano Market Brief is published weekly. Read past issues in Market Insights & Newsletters, browse explainers in Learn with Mugano, or see our Security & Compliance page. Existing clients can sign in.

Important risk information

Risk warning: Don't invest unless you're prepared to lose all the money you invest. Crypto-assets are high-risk and highly volatile: their value can go down as well as up, you may lose your entire investment, and you should not expect protection from compensation or ombudsman schemes if something goes wrong. Past performance is not a reliable indicator of future results. This newsletter is for information and education only, it is not investment, financial, tax, or legal advice, nor a recommendation or solicitation to buy or sell any asset. We are not authorised to provide financial advice and nothing here takes account of your personal circumstances. Always do your own research and consult an independent, licensed financial adviser before investing. Profits may be subject to tax depending on your circumstances.

United Kingdom: Don’t invest unless you’re prepared to lose all the money you invest. Cryptoassets are high-risk, can fall in value quickly, and are not covered by FSCS protection. This newsletter is general information only and is not investment advice.

EU / EEA: Crypto-assets are high-risk and you could lose all the money you invest. They may be volatile, illiquid, and not covered by deposit guarantee or investor compensation schemes. This communication is for general information only and is not investment advice.

Australia: Crypto-assets are high-risk and highly volatile, and you may lose your entire investment. This newsletter is general information only and does not take into account your objectives, financial situation, or needs. Seek independent advice before acting.

United States: Digital assets are speculative and involve a high degree of risk, including possible total loss. Prices can be volatile, and platforms may not provide FDIC, SIPC, or other protections. This communication is for general information only and is not investment, legal, or tax advice.