General commentary for a general audience, not personalised advice. Price is the US close on 9 Oct 2026 ($84.51, market cap about $21.6B, StockAnalysis). Company figures are from Circle's second quarter 2026 10-Q and results release (5 Aug 2026, re-checked on SEC EDGAR 12 Oct 2026). Stablecoin supply is from DefiLlama, checked 12 Oct 2026. All scenario assumptions below are ours and are labelled as such. Circle reports Q3 on 4 Nov, so expect some of this to move. This is a companion to our Circle vs Coinbase note, not a replacement.
Here is the tension. Almost everyone agrees stablecoins will be much bigger in five years. Far fewer people have done the arithmetic on how much of that growth actually reaches Circle's shareholders after interest rates, partners and competitors take their share. The story is exciting. The economics are narrower than the story.
From poker chip to plumbing
A stablecoin is a token that is always worth one dollar because a real dollar, or a Treasury bill, sits in a reserve behind it. For most of their life stablecoins were poker chips: traders parked money in them between crypto bets. That is still a big part of the demand today.
What is changing is who else uses them. Companies are using them to move money across borders in minutes instead of days. Remittance apps use them to cut fees. Funds use them as cash inside tokenised versions of Treasury funds and other assets. Payment giants are building on them. Stripe closed its $1.1B purchase of the stablecoin platform Bridge on 4 Feb 2025 (Stripe, CNBC). Visa launched USDC settlement for US banks on 16 Dec 2025, when its stablecoin settlement had passed a $3.5B annualised run rate (Visa press release).
The newest use is the one that gets the most attention, AI agents paying for things. An AI agent is software that acts on your behalf, for example booking, buying data or calling another service. Agents cannot open bank accounts or sign card agreements, but they can hold a wallet. Coinbase's x402 standard, launched with AWS, Anthropic, Circle and NEAR as collaborators, lets a website ask for a payment inside an ordinary web request and lets an agent pay it in USDC in seconds. Circle says x402 processed more than $100M of payments in its first few months. On 29 Apr 2026 Circle switched on Nanopayments, USDC payments as small as $0.000001 that cards cannot do economically, across 11 blockchains (Circle blog). On 16 Sep 2026 it launched the public mainnet of Arc, its own blockchain built around USDC, with Visa among its design partners, and said USDC accounts for 98.8% of agent driven transaction volume (Circle press release, 16 Sep 2026, Circle's own figure). If machines start buying from machines at scale, the natural currency is a programmable dollar, and USDC is the regulated one most of these projects default to.
We think this is real, but honestly early. Agent payments today are tiny next to trading balances. They matter for the bull case, not the next two quarters.
How Circle actually makes money
Circle earns interest on the reserve behind USDC and keeps whatever is left after paying partners. In Q2 2026, reserve income was $668M out of $701M of total revenue, so 95% of the business is interest. USDC in circulation ended the quarter at $73.3B and the reserve earned about 3.5% (Circle calls this the reserve return rate).
Then come the partners. Circle paid $410M in distribution and transaction costs in the quarter, about 59% of revenue, mostly to the exchanges and apps that hold or push USDC, with Coinbase the largest by far. That leaves $289M, which Circle calls revenue less distribution costs. Its own staff, compliance and technology come out of that. Operating expenses were $254M in the quarter, about $1.0B a year at that pace, and over the past twelve months Circle's operating income was $216M (S&P Global via StockAnalysis, 12 Oct 2026).
So there are really only four numbers that matter for the long run: how many USDC exist, what the reserve earns, what share goes to distributors, and how much other revenue Circle can build from services like Arc and its payments network. Everything below is those four dials.
The rate dial comes first
Before we get to growth, one point that stablecoin enthusiasts tend to skip. Circle's 10-Q says that if rates were 100 basis points (one percentage point) higher than June's average for a year, with supply unchanged, reserve income would rise by about $737M, distribution costs by about $360M, and Circle would keep about $377M more before tax. The same works in reverse. That $377M is larger than Circle's entire operating income over the past year.
Right now the dial is helping. The Fed raised rates a quarter point on 16 Sep 2026 to a range of 3.75% to 4.00% (FOMC statement). But in every scenario below, our reserve yield assumption does as much work as our growth assumption. If you own Circle, you own a view on short rates whether you meant to or not.
Three futures for stablecoins
For the size of the whole market we lean on outside forecasts. Total stablecoins are about $313B today, with Tether's USDT at $184B and USDC at $73B, roughly a 23% share (DefiLlama, 12 Oct 2026). Citi's Stablecoins 2030 report puts the market at $0.9T in its bear case, $1.9T in its base case and $4.0T in its bull case by 2030. Standard Chartered reaffirmed in February 2026 that it expects $2T by the end of 2028, while noting growth had stalled around $300B. The US Treasury Borrowing Advisory Committee (TBAC), in its report to the Treasury Secretary on 30 Apr 2025, noted estimates of about $2T of stablecoins by 2028, up from about $234B at the time. These are forecasts from interested parties in a fast moving market, not facts, and the stall since early 2026 is a reminder.
Our three scenarios, all ours, for 2029:
Bear: still mostly a trading chip. Crypto stays weak, the GENIUS Act rules squeeze rewards that keep balances parked on exchanges, and banks issue their own tokens for corporate clients. USDC grows only to about $90B. The Fed is cutting by then and the reserve earns 2.5%. Distributors take 62% of reserve income because Circle has to pay up to keep shelf space. Other revenue is about $0.2B. Revenue is about $2.5B and operating profit is about zero. At around 2 times sales plus Circle's $1.73B of cash (it has no debt, 10-Q), that is about $24 a share in 2029, or roughly $17 in today's money.
Base: payments plumbing grows, rates normalise. Our own assumption is that the total market reaches roughly $0.9T by 2029, a figure we interpolated on the way to Citi's 2030 base case rather than one anyone published, and USDC keeps a bit over a fifth of it, about $200B. The reserve earns 3.25%, distributors take 60%, and Arc, the payments network and other services bring in $0.5B. Revenue is about $7.0B, operating profit about $1.6B, and EPS (earnings per share) about $4.30 on our assumption of 280M shares, which allows for roughly 11% more dilution from the 253M Class A and Class B shares outstanding at 30 Jun 2026 (10-Q). At 20 times earnings, discounted back at 10% a year, that is worth about $64 today.
Bull: the dollar's default rail, agents included. Standard Chartered's $2T arrives on schedule, USDC holds about 22% for roughly $450B, payments and agent commerce make USDC the default for regulated money, and Circle's own platforms carry more of the supply so distributors' share falls to 55%. Rates hold and the reserve earns 3.5%. Other revenue reaches $1.5B. Revenue is about $17B, operating profit about $6.6B and EPS about $17.90. At 22 times earnings, discounted, that is worth about $290 today.
Every row is our assumption, not a forecast from Circle or anyone else. Operating costs are also ours (about $1.1B, $1.5B and $2.0B a year), scaled up from today's run rate of about $1.0B. Tax is 24%, and values are discounted at 10% a year from the end of 2029. Base and bull ignore today's cash, which would add about $6 a share. Probability weighted, the three come to about $95, a little above today's price, but with a very wide spread.
What the price is paying for
A reverse DCF works backwards from today's share price to the future the market must be assuming. Keeping our base case settings for yield, payout, costs and multiple, $84.51 only adds up if USDC reaches about $240B by 2029, roughly 3.3 times today's supply. That is more than our base case and less than half the way to our bull case.
Now hold supply at our base case $200B and cut the reserve yield to 2.25%, which is what a full easing cycle would look like. Value drops to about $32. That is the rate sensitivity from the 10-Q showing up in the share price. The market is paying for meaningful growth and for rates staying roughly where they are, both at once.
The competition and the payout problem
Tether is the giant. USDT is about two and a half times the size of USDC and dominates offshore trading and emerging market savings. Circle's edge is regulation: it is the one US and European institutions can hold without a compliance headache. Under GENIUS, that edge matters more inside the US, but on 27 Jan 2026 Tether launched USA₮, a US regulated coin issued by Anchorage Digital Bank under the GENIUS framework (Tether, CoinDesk), so the moat is a head start, not a wall.
The bigger long run threat may be the companies that own customers. PayPal's PYUSD is about $2.9B (DefiLlama). Stripe owns Bridge, which lets businesses issue their own branded stablecoins, and incubated Tempo, a payments blockchain that went live on 18 Mar 2026 with a machine payments protocol for agents co-written with Stripe (Tempo, The Block). That is a direct rival to Arc and x402. Banks are exploring tokenised deposits. Each of these keeps the reserve interest for itself instead of sending it to Circle. That is why the distributor line is the number we watch most closely. Circle's partners already take about 59% of total revenue (61% of reserve income), and the renewed Coinbase agreement keeps its split in place. Every time a big platform decides it can issue its own coin, it either leaves or asks for more.
The GENIUS Act clock
The GENIUS Act, the US stablecoin law signed on 18 Jul 2025, takes effect on 18 Jan 2027 (Treasury, 17 Aug 2026). It requires full reserves in cash and short Treasuries, which Circle already holds, and it bans issuers from paying holders interest just for holding a coin. The OCC's March 2026 proposal would presume indirect payments through partners are banned too, and its final rule is expected by November. That is the near term swing factor. Clarity helps Circle win institutional balances. A strict reading on rewards could shrink the balances that sit on exchanges today.
The bear case, taken seriously
The strongest bear argument is that Circle is a regulated utility that pays most of its income to whoever owns the customer, and that the stablecoin boom will mostly enrich Tether, Coinbase, Stripe and the banks. USDC hit $79.66B on 17 Mar 2026 (stables.cool) and has slipped to about $73B since. The Fed could be cutting by 2028, and lower rates hurt directly. And the stock is not cheap on today's numbers: consensus has 2026 adjusted EPS falling 51% to $1.14 (StockAnalysis), so you are paying about 74 times this year's earnings for a story about 2029. Analysts' low target is $37. We do not think this is the most likely outcome, but it is a perfectly ordinary one.
What would change our mind
We would get more constructive if USDC climbs back above its March peak while rates hold, if the distributor share drops noticeably as more USDC sits on Circle's own platform ($12.4B at the end of Q2, up 106% in a year), or if Arc and agent payments start showing up as real revenue rather than announcements.
We would get more cautious if the final OCC rule kills partner paid rewards, if a large platform such as Stripe or a big bank pulls volume into its own coin, or if the Fed signals a return to cutting.
Where the terminal sits
The TraderMagz terminal reads Circle as "Tradable, not core" as of 12 Oct 2026 08:53 UTC (10:53 SAST), with conviction 43 and timing 57. It is outperforming a lagging Financials sector. The fundamentals leg dates from 1 Oct 2026 and the terminal flags the data as stale, so treat this as a cross check.
What we are watching
- OCC final rule, expected by November: how it treats rewards paid through partners.
- Circle Q3 on 4 Nov: USDC supply, reserve yield after the September hike, distribution costs as a share of revenue, and any revenue line from Arc or payments.
- Total stablecoin supply: whether the market breaks out of its roughly $300B range.
- Fed path: every 100 basis points is worth roughly $377M a year to Circle before tax.
Bottom line
Stablecoins becoming the internet's dollar, for people, companies and eventually AI agents, is one of the better structural stories in finance, and Circle is the purest listed way to own it. But at $84.51 the market is already paying for USDC to more than triple by 2029 with rates holding up, which is above our base case. Our probability weighted value of about $95 says the odds lean slightly in your favour, yet most of that comes from a 20% bull case, and a full rate cutting cycle could roughly halve the base case. This is a stock for people who believe in the bull case and can live with rate risk, sized like the option it is.