In the U.S. preparation time is shrinking;
in Korea the market is moving ahead of the law

On July 18, 2026, U.S. regulators had adopted no final GENIUS Act rules at all^[1]. The GENIUS Act, enacted on July 18, 2025, is the federal statute governing payment stablecoins in the United States. It required implementing regulations within one year of enactment. The agencies named are the Office of the Comptroller of the Currency (OCC), the Federal Reserve, the Federal Deposit Insurance Corporation (FDIC), the National Credit Union Administration (NCUA) and the Treasury. The first four are designated primary federal payment stablecoin regulators, which determines who supervises each issuer. One law firm counted 11 proposed rules and zero final rules as of July 17, 2026^[6].

The statute sets no penalty and no fallback schedule for missing the deadline^[1], and the law does not lapse. The effective date provision, however, still stands. Section 20 sets the effective date as the earlier of January 18, 2027, which is 18 months after enactment, or 120 days after the primary regulators finalize their rules^[4][5].

That yields one date worth putting on a calendar. If the rules are finalized after September 20, 2026, the 120-day count lands beyond January 18, 2027, so the 120-day route can no longer pull the effective date forward (our estimate)^[1][5]. Late rules do not push the effective date back. What shrinks instead is the time firms have to apply for a license and build their reserve, redemption and disclosure systems.

The Federal Reserve issued no rule of its own and only joined a shared one

Of the four primary regulators, only the Federal Reserve has yet to produce its own rule. It is the primary regulator for state member banks, yet it has not proposed GENIUS Act implementing regulations^[6]. Instead it appears as one of five issuing agencies on the joint customer identification proposal published in the Federal Register on June 22, 2026 (91 FR 37234)^[2]. That rule treats permitted payment stablecoin issuers as financial institutions under the Bank Secrecy Act and requires them to maintain a customer identification program^[2].

The same proposal contains two figures worth noting. The Treasury and the four agencies estimate roughly 50 permitted issuers over the first three years the Act is in effect, of which about 60% would be subsidiaries of insured depository institutions^[2]. In other words, regulators expect an initial licensed market of that size, with depository-institution subsidiaries accounting for a large share. The customer identification rule itself proposes a compliance date 12 months after the final rule is issued^[2].

The OCC announced its proposal on February 25, 2026, and it appeared in the Federal Register on March 2^[1][4]. The table uses publication dates throughout; the timeline chart uses agency announcement dates. Three anti-money-laundering proposals published between April and July 2026 implement the anti-money-laundering statutes rather than the GENIUS Act and are excluded from the table^[3][6]. As of Aug 12, 2026.

The previous chapter showed that the effective date is fixed while the rules are late. So what happened in the meantime to the firms those rules will govern? Circulation, return rates, transaction activity and the competitive field all keep moving while issuers wait for the rules to set their operating terms. Circle's second-quarter 2026 results, from one of the largest issuers that will fall under the regime, show that shift in one place.

Circle reported second-quarter results on August 5, 2026^[8]. USDC in circulation stood at $73.3 billion at quarter end, up 19% year over year. Onchain transaction volume reached $14.8 trillion, up 151%. Total revenue and reserve income, by contrast, came to $701 million, up only 7%. Reserve income was $668 million, up 5%, as a 66 basis point decline in the reserve return rate offset a 25% rise in average USDC in circulation to $76.5 billion^[8].

What has shifted is not the reliance on reserve income itself. Reserve income still accounts for 95.3% of total revenue (our estimate)^[8]. Dividing reserve income by average circulation and annualizing puts the reserve return rate at roughly 3.5% (our estimate)^[8]. What changed is that rate and the quarter-end balance. Circulation fell from $77.0 billion at the end of the first quarter to $73.3 billion, a decline of 4.8% (our estimate)^[8][9]. With the Federal Open Market Committee holding the policy rate at 3.50-3.75% on July 29, the drop in the return rate stays put unless rates rise^[13].

Efforts to widen the revenue base are visible. Other revenue, excluding reserve income, was $34 million, up 41% year over year but down 19% from the previous quarter^[12]. Circle raised its 2026 other revenue guidance from $150-170 million to $310-330 million. The increase includes token presale revenue from Arc, its own blockchain, which is scheduled to launch on September 16^[12]. Whether that becomes recurring service revenue will be clear only after launch.

Analyst downgrades and the company's own risk disclosure point to the same place

On July 30, 2026, Bernstein cut its 2028 forecast for USDC in circulation from $290 billion to $170 billion, a reduction of 41%^[11]. On August 3, Morgan Stanley moved to underweight and lowered its circulation assumptions by about 33% for 2027 and about 44% for 2028^[10]. Both cite the same reason: competing products are drawing money away. The law bars USDC from paying interest to holders, while tokenized money market funds and tokenized deposits can offer a yield^[10].

Circle records the same risk itself. The risk factors in its second-quarter disclosure include the rise of yield-bearing digital assets, tokenized money market funds among them, which could reduce demand for stablecoins and shrink circulation^[8]. The analysts' diagnosis and the company's own assessment converge.

Circle already holds a federal charter separate from the GENIUS Act track

Circle received final approval from the OCC to establish Circle National Trust, a national trust bank^[8]. The charter permits federally regulated digital asset custody and opens a path for Circle to manage USDC reserves directly^[8]. The New York State Department of Financial Services separately approved a limited purpose trust company focused on digital assets^[8].

This route is distinct from the general issuer licence the GENIUS Act rules will define. A national trust charter serves custody and reserve management; it does not substitute for the permitted issuer requirements the rules will set. It also means the delay does not weigh equally on every firm. A company already inside the federal supervisory perimeter carries a different preparation burden from one that is not.

The macro estimates have not moved since February

The Treasury bill demand estimate covered in the previous issue has not been updated. In February 2026, Standard Chartered projected that stablecoins would add $800 billion to $1 trillion in short-term Treasury demand. No revision has been obtained since (secondary citation, primary report not obtained)^[15][16]. That estimate rests on an assumption that stablecoin market capitalization reaches $2 trillion by 2028. Two trillion dollars is 6.64 times the $301.2 billion retrieved on August 11, 2026 (our estimate)^[14].

Total market capitalization has hovered in the $300 billion range since October 2025. Against DefiLlama's July 19 figure of $310.1 billion, the August 11 reading is about 2.9% lower, and the same series shows a 30-day change of minus 0.94% (our estimate)^[14][18]. On the same retrieval, USDT stood at $183.1 billion, 2.53 times USDC, with issuance shares of 60.78% for USDT and 24.01% for USDC^[14]. Quarterly series for Tether were not obtained, so this report does not rule on whether the shift in conditions around reserve-based revenue is common to the industry.

The Federal Reserve Bank of Kansas City offers a different angle. It finds that stablecoins may raise Treasury demand only by reducing demand for other assets^[17]. If the effect is a reallocation rather than net new demand, the macro impact is smaller than the headline estimates suggest.

Now to Korea. The United States passed a law and stalled at the rulemaking stage; Korea has stalled one step earlier, at the statute itself. What is stuck in Korea, though, is the review rather than the bills. Ten bills on digital assets and stablecoins are pending before the National Assembly, beginning with the measure sponsored by Representative Min Byoung-dug on June 11, 2025^[21]. Some were taken up by the National Policy Committee in August 2025 and referred to its legislative review subcommittee, but the consolidated review that would merge them into one bill has not been held^[24].

The government bill has not been submitted either. At the committee's July 29 briefing, Financial Services Commission Chairman Lee Eok-won said the bill was ready but some differences remained^[21]. In its second-half 2026 growth strategy, announced on July 14, the government set legislation within the year as a goal^[26]. The intent is confirmed repeatedly; the text is not yet public.

The schedule has slipped as well. An extraordinary session opened in August, but lawmakers expect substantive review to move to the September regular session^[19]. Party and government consultations require the Democratic Party to seat new leadership and a policy committee after its August 17 national convention^[19]. The digital asset task force has finished its work, and whether it will be revived is undecided^[19]. The National Policy Committee did agree on July 28 to convene each of its two legislative review subcommittees twice a month, so the review framework is in place^[21]. With the October audit of government agencies and budget review in November and December to follow, observers say passage within the year becomes difficult unless a unified bill emerges in September^[20].

Banks at 50% with fintechs at 34% is a political proposal, not the government bill

The compromise under discussion would give a bank consortium 50% plus one share, with fintech firms holding about 34%. The proposal comes from Representative Ahn Do-geol, who served as secretary of the Democratic Party's digital asset task force in the first half of the 22nd National Assembly^[22]. The 34% figure has a basis. Under the Commercial Act, holding more than one third of total shares outstanding, or 33.4%, allows a shareholder to block special resolutions such as amendments to the articles of incorporation or a merger. Fintechs would need at least that much to have a say in management^[22]. In public remarks on June 23, Ahn also described a division of roles in which a bank consortium takes the position of nominal controlling shareholder and backstop while fintechs lead day-to-day management^[23].

This structure is not settled as the government bill. The Financial Services Commission has repeatedly said that the issuer structure, shareholder composition and other core elements of the second-stage legislation remain undecided^[27]. The 34% figure is proposed as a floor for fintech ownership, not a ceiling^[22].

Exchange ownership is a separate flashpoint. Through the first half of the year, no agreement was reached on whether to apply a 15-20% ownership cap uniformly to controlling shareholders of digital asset exchanges including Dunamu, Bithumb, Coinone, Korbit and Streami^[22]. Ahn favours adopting the cap with a grace period before it takes effect^[22].

Sources: Edaily[21][22], Blockmedia[19][20], E-Focus[24], Korea Policy Briefing[27]. As of Aug 12, 2026.

Chapter 02 showed how conditions changed on the issuer side in the United States while the rules were pending. This chapter looks at how conditions changed on the user side in Korea while the legislation was pending. Domestic holdings have tilted toward the dollar. Bank of Korea data submitted to the National Assembly shows a sharp rise in stablecoin holdings at the five largest domestic exchanges. Holdings grew 6.9 times, from KRW 88.5 billion at the end of July 2024 to KRW 607.1 billion at the end of February 2026 (our estimate)^[28]. The peak, however, was KRW 872.3 billion at the end of December 2025, leaving the latest reading 30.4% below it (our estimate)^[28]. In absolute terms, KRW 607.1 billion is roughly 1% of the KRW 60.6 trillion in domestic crypto holdings at the same date (our estimate)^[28].

Everything else shrank over the same period. Domestic crypto holdings fell from KRW 121.8 trillion at the end of January 2025 to KRW 60.6 trillion at the end of February 2026, less than half^[28]. Won deposits held at exchanges also fell, from KRW 10.7 trillion at the end of December 2024 to KRW 7.8 trillion at the end of February 2026^[28]. Trading points the same way. On Kaiko data, the combined share of USDT and USDC in turnover at Korea's four largest exchanges rose markedly year over year. Cumulative turnover for January through May 2026 was about $53.8 billion^[29].

The Kakao, Toss and Circle payment agreements remain non-binding

A separate development sits alongside the shift in user assets. In its second-quarter results, Circle said the Kakao group had begun exploring blockchain payment infrastructure and USDC integration in Korea^[8]. On July 23, Circle signed memoranda of understanding with Kakao, Kakao Pay and Kakao Bank, and separately with Viva Republica and Toss Bank; in April it signed with Upbit and Bithumb^[30][31]. Circle says it has no plans to issue a won stablecoin itself and will support domestic efforts by providing technology^[31].

All five are memoranda of understanding rather than binding contracts, and no launch timetable has been disclosed^[30][31]. They are not evidence that Korea's payment rails have shifted to dollar stablecoins.

One fact runs through all four chapters. Conditions for markets and firms are already changing before the framework is complete. The United States wrote the law first and has not finalized the rules; in Korea the statute itself waits for a consolidated review in subcommittee. U.S. firms have less time to prepare before the effective date. In Korea, user assets and payment infrastructure discussions moved first while second-stage rules on issuance, circulation and disclosure remain unfinished.

The evidence leans one way. The missed deadline and the persistence of proposals are confirmed by the Federal Register text, a law firm's tally and independent reporting from several outlets^[1][2][6]. The change in issuer conditions rests on Circle's second-quarter disclosure, a primary source^[8]. Korea's unsubmitted government bill and slipped schedule are corroborated by the FSC chairman's answers before the National Assembly and by political reporting^[19][21]. The shift in domestic holdings rests on Bank of Korea data submitted to the National Assembly^[28].

Three areas remain beyond judgment. The shift in conditions around reserve-based revenue was confirmed only for USDC and Circle, and quarterly series for Tether were not obtained^[14]. The rise in domestic dollar stablecoin holdings also involves exchange rate effects and is not explained by legislative delay alone^[28][29]. No data directly measures how much preparation time U.S. firms have actually lost.

1. Limitations

• Scope of the issuer assessment: the shift in conditions around reserve-based revenue was confirmed for USDC and Circle only. Quarterly circulation series for Tether were not obtained, so we do not rule on whether this is common to the industry.
• Series with different start dates: the earliest publicly available data point is end-July 2024 for stablecoin holdings, end-December 2024 for won deposits and end-January 2025 for total crypto holdings. Comparison over an identical window was not possible, so each series is shown over the range available.
• Lag in the Korean figures: the latest published value for domestic user assets is as of end-February 2026, five months before publication.
• Path to the Bank of Korea data: these figures come from data the Bank of Korea submitted to the National Assembly, as reported by the press. The submitted document itself was not obtained.
• Choice of aggregator: stablecoin market capitalization and issuer-level size use a single DefiLlama retrieval dated August 11, 2026. The dashboard renders in JavaScript and cannot be collected automatically, so it was retrieved manually. Aggregators covering a different set of coins may report different values.
• Effective date trigger: Section 20 of the GENIUS Act keys off the date the primary regulators issue final regulations but does not specify whether all four or only some must act. This report calculates on the basis of all four; the actual reading may differ.
• Causation not established: we could not establish causation between the ban on paying yield and the decline in USDC circulation. Crypto markets contracted broadly over the same period and no data separates the two factors.
• Preparation time not measured: no data directly measures how much preparation time U.S. firms have lost. The statement is derived from the structure of a fixed effective date combined with delayed rules.
• Primary report not obtained: the Standard Chartered Treasury demand estimate is cited through press reports because the original was not available.

2. Formulas for Our Own Estimates

Every figure marked as our estimate in the main text is set out here with its formula and basis.

3. Methodology Note (standing across issues)

• Source hierarchy: company and regulator filings, then primary data such as aggregator dashboards, then specialist media, then syndicated outlets, then community material.
• Primary versus secondary: only sources whose original text we read are marked primary. Where the original was not obtained, the main text says so and flags the secondary citation.
• As-of and retrieval dates: fixed-point figures carry an as-of date; dashboard values that move in real time carry a retrieval date.
• Conflicting figures: both are presented. The choice of a reference value rests on whether the calculation scope is disclosed, whether a retrieval point can be stated and how finely the items are separated, not on the standing of the outlet. The reason is given in the main text.
• Metric layers: company revenue, issuance outstanding, trading volume and assets held are kept distinct and are not compared as shares of a single whole.
• Layers of legal instruments: statutes, enforcement decrees, implementing regulations, public notices and guidelines differ in who issues them and how binding they are, so they are named individually rather than lumped together.
• Procedural status: the status of legislation and licensing is described using confirmed stage names such as filing, referral, review and passage, together with the actor involved.
• Our estimates: formula, as-of date and limitations are disclosed one by one in Appendix 2.
• Conflict of interest: the same disclosure appears in a fixed position at the top of the document in every issue.