Last June, Circle went public at $31 per share. Two weeks later, the stock price reached $299. Then, it fell back to $50, a drop of more than 80%. Then, after the earnings report came out this February, the stock price doubled in two weeks, reaching $111 today.

During the same period when it doubled, Bitcoin fell by 40%. Circle's stock price lost its connection with the crypto market.

"You are seeing that kind of decoupling."

The drop to $50 is not hard to understand within the industry.

In September and October 2025, the Federal Reserve lowered the benchmark interest rate by 25 basis points twice, bringing it down to 3.75%. Reserve yields subsequently declined; Q3 data shows that the USDC unit reserve yield fell by 96 basis points year-on-year.

How much money do those 96 basis points represent? Circle calculated in its prospectus that every 100 basis point cut in interest rates by the Fed results in an annualized loss of approximately $618 million in interest income. Half of this loss is offset by a corresponding decrease in allocation costs, resulting in a net loss of approximately $300 million. However, this is only a static calculation, assuming the size of USDC remains unchanged. In 2025, the Fed cut rates by a total of 75 basis points; this single variable of interest rates alone will take nearly $200 million from Circle's annual revenue.

That's not all. Circle and Coinbase have a revenue-sharing agreement. All reserve revenue generated by USDC held on the Coinbase platform goes to Coinbase, while revenue outside the platform is split 50/50 between the two companies. In Q4, total reserve revenue was $733 million, allocation costs were $461 million, and Circle actually retained net reserve revenue of $273 million. Non-interest income was $37 million, less than 5% of total revenue.

This structure is the fundamental reason why Circle's stock price fell from $299 to $50. Interest rate cuts compressed reserve yields, Coinbase's revenue sharing structure was fixed, and Circle, sandwiched in the middle, had a clear ceiling on how much revenue it could collect and a clear floor on how much profit it could retain.

However, when the financial report was released on February 25, Circle's stock price rose by 35% in a single day.

EPS was $0.43, exceeding analysts' expectations of $0.16. This wasn't just a slight overshoot; it was a major disappointment. But what drove the repricing wasn't just the numbers in this quarterly report, but a larger structural fact that this report simply revealed to the market for the first time.

In 2025, the entire cryptocurrency market capitalization fell by more than 40% from its peak. During this period, the circulating supply of USDC increased by 72% to $75.3 billion, a record high. The total market capitalization of stablecoins also surpassed $314 billion during the same period, also a record high.

It's not a tailwind in a bull market, but a headwind in a bear market.

This event is fundamental to Circle's pricing logic. Previously, the market's valuation framework for CRCL treated it as a beneficiary of the crypto cycle: in a bull market, people use USDC for transactions, leading to increased scale; in a bear market, reduced on-chain activity leads to decreased scale. Coinbase's logic is similar: transaction volume is lifeblood, and in a bear market, fee revenue plummets.

However, data from 2025 refutes this framework. USDC's growth did not stop during the bear market; in fact, it accelerated.

Circle CEO Allaire said during the earnings call, "You're seeing that kind of decoupling." He was referring to the decoupling between Bitcoin and stablecoins. But what he didn't finish was that the use case for stablecoins is shifting from a unit of account for crypto transactions to a settlement infrastructure for global payments.

The driving force is no longer speculative trading demand, but rather the entry of a group of participants who have never appeared in the crypto space before. Visa announced the expansion of USDC settlement, allowing US Visa card issuers to complete settlements with USDC outside of normal banking hours. Mastercard followed suit, and JPMorgan Chase launched several USDC-related products last year. Intuit announced a partnership with Circle to bring low-cost programmable payments to its tens of millions of corporate and individual customers, and Polymarket completed a large-scale migration of USDC as its core settlement asset.

This isn't about crypto native users depositing and withdrawing money. This is about traditional financial institutions embedding USDC into their payment and settlement pipelines. These two use cases correspond to completely different valuation logics. The former follows the crypto cycle, while the latter follows the global payment volume. The global cross-border payment market is approximately $150 trillion annually, and USDC's current quarterly on-chain transaction volume is $11.9 trillion, a year-on-year increase of 247%. These two figures cannot be directly compared, but the market is beginning to use this second framework to price Circle.

The GENIUS Act, and a mere admission ticket.

A widely circulated saying on X is: "If your argument is that stablecoins are eating up global payments, then CRCL is the most direct bet. COIN is a hybrid entity that incidentally takes a cut from USDC; the two are different instruments addressing different issues."

This explains why $CRCL was able to independently drive a price surge while Coinbase's stock price remained flat. Coinbase operates an exchange, wallets, the Base Chain, and provides institutional custody services; USDC is just one of its many business lines. Circle, on the other hand, focuses on only one thing: issuing and circulating USDC. If you want to bet on the stablecoin sector itself, there's only one pure asset available on the market.

This logic became clearer after the passage of the GENIUS Act.

In July 2025, the bill was enacted, establishing the first federal regulatory framework for stablecoins. It requires compliant issuers to hold 100% cash or short-term Treasury bond reserves and to undergo regular audits. On the day the GENIUS bill was passed, $CRCL surged 34% in a single day. The market understood this signal. This wasn't just a compliance benefit; it was a regulatory barrier drawn between USDC and USDT, a line that Tether cannot cross in the short term.

JPMorgan's data confirms this assessment: after the bill's passage, the overall stablecoin market grew by 19%, with USDC's market share rising from 24% at the beginning of the year to 25.5%, while Tether's fell from 67.5% to 60.4%. The on-chain transaction volume figures are even more telling: USDC surpassed USDT in Q4, capturing approximately 50% of the stablecoin's on-chain transaction volume—the first time in many years that Tether has been overtaken in this metric.

But Tether didn't give up. After the GENIUS Act passed, Tether launched USAT, partnering with Anchorage Digital and Cantor Fitzgerald, specifically designed with reserves according to GENIUS standards. USAT's CEO, Bo Hines, is a former White House crypto advisor. Currently, USAT's circulating supply is approximately $20 million, negligible compared to USDC's $75.3 billion. However, Tether possesses the world's largest stablecoin user network, and Cantor Fitzgerald brings Wall Street connections—a combination not to be underestimated.

Meanwhile, a number of names never before seen in the stablecoin arena have entered the fray. Fidelity launched FIDD, running on Ethereum with a 100% GENIUS Standard Reserve, targeting institutional and retail markets. Robinhood and Revolut are reportedly developing their own stablecoins. JPMorgan Chase and US Bancorp are expanding their stablecoin initiatives. Stripe, after acquiring Bridge, has embedded its stablecoin settlement pipeline into its $1.9 trillion annual payment flow. Treasury Secretary Bessent stated that the US stablecoin market could reach $3.7 trillion by the end of this decade. USDC's current $75.3 billion is less than 3% of that figure.

The GENIUS Act didn't just open the door for Circle; it opened the door for the entire traditional financial system. The first-mover advantage is real: 30 blockchain-native supports, deep integration with Visa and JPMorgan Chase, and years of accumulated enterprise API infrastructure—these are things new entrants can't replicate in a year or two. Bernstein gives a target price of $190, citing regulatory moats and competitive barriers in its technology stack.

However, the depth of Circle's competitive advantage remains unknown until the real stress test arrives. Another issue rarely discussed openly within the industry is the fixed term of Circle's revenue-sharing agreement with Coinbase. In the next round of negotiations, Coinbase's leverage will not be less than it is now; USDC's share on its platform has increased from 5% in 2022 to 22% currently. The outcome of the negotiations will directly impact how much of the USDC growth Circle can actually retain.

$23 billion, betting on a story that hasn't happened yet.

Circle's current valuation is largely based on a story that hasn't even happened yet.

Allaire devoted considerable time during the earnings call to discussing the payment needs of AI agents. When performing autonomous tasks, AI agents require small, frequent, cross-time zone payments, involving API calls, purchasing computing power, and completing cross-border settlements. Allaire calls this scenario the "machine economy," arguing that when the number of AI agents exceeds the number of human users, the primary users of payment infrastructure will no longer be humans, but machines.

These demands, when placed within the traditional payment system, present friction at every step. Credit cards have limited business hours, require manual authorization, and their minimum fee structure makes payments under $0.01 economically unfeasible. Stripe charges a minimum of $0.30 plus 2.9% per transaction, Visa and Mastercard's cross-border fees average 1.5%-3%, and bank wire transfers do not operate on weekends.

USDC doesn't have these technical limitations; it operates 24/7, settles on-chain, and costs less than $0.001 per transaction on high-speed chains like Solana, while Arc aims for $0.00001. Circle has specifically developed payment infrastructure for AI agents, and the Arc testnet is already live. This isn't just a "slightly cheaper" improvement; it represents an order-of-magnitude difference in the entire cost structure.

Benchmark-StoneX analyst Mark Palmer put it bluntly: "AI agents need programmable currency that can be directly embedded into software workflows, without long settlement windows," while card organizations' infrastructure is designed for human checkout processes, not for machines.

How real this demand is can be seen from the speed of action at the protocol layer.

In May 2025, Coinbase launched the x402 protocol, using the long-dormant HTTP 402 status code to enable automatic payments by its AI Agent. This allows servers to directly request USDC payments before responding to requests, eliminating the need for manual authorization. Five months later, x402 processed over 100 million payments. Google launched AP2 (Agent Payment Protocol), and OpenAI internally tested "Instant Checkout" in ChatGPT, with the underlying settlement layer being a combination of Stripe and stablecoin tracks. These are not white papers; they are infrastructures already running in production environments.

Visa's data provides a benchmark for scale. In November 2025, Visa's monthly annualized transaction volume settled via stablecoins was approximately $3.5 billion, rising to $4.5 billion annualized by January 2026. Compared to Visa's total annual transaction volume of approximately $16 trillion, this figure is almost negligible. However, the change in direction is more noteworthy than the absolute value; Visa itself is expanding this pipeline. Coinbase CEO Brian Armstrong also made the same assessment in early March: "Soon, the number of AI agents initiating transactions will surpass that of humans."

The data clearly shows the gap between narrative and reality.

x402's total transaction volume over the past 30 days was $24 million, with 94,000 buyers and 22,000 sellers participating. The global e-commerce market size during the same period was estimated at $6.88 trillion. McKinsey estimates that the current real-world payment use of stablecoins is approximately $390 billion per year, of which about $226 billion is B2B, with retail accounting for even less. ECB data shows that organic retail transfers account for approximately 0.5% of total stablecoin traffic.

Circle reported a net loss of $70 million for the full year of 2025. The Mainnet is planned for launch in 2026 and is currently still a testnet. AI and non-interest income combined accounted for less than 5% of total revenue for the year.

Gartner predicts the AI ​​agent economy will reach $30 trillion by 2030, while Bessent predicts the stablecoin market will reach $3.7 trillion by the end of this decade. If these figures are accurate, Circle's current circulating USDC of $75.3 billion is indeed just the beginning. However, the journey from $24 million x 402 monthly transaction volume to a $30 trillion agent economy is a path no one has ever walked before.

The $23 billion valuation is betting that this journey will be completed.

Circle went public last June at $31, rose to $299 two weeks later, then fell back to $50, and has now doubled again. One question along this curve has never been truly answered: What kind of company is Circle ultimately?

Is it an interest rate arbitrage business, a compliant stablecoin infrastructure, or a settlement layer for the AI ​​economy? Allaire says, "You're seeing that kind of decoupling," but where it will decouple to is another question.