---
ticker: CRCL
company: Circle Internet Group, Inc.
analyst: customer-analyst
generated: 2026-05-12
side: long (long-term compounder lens, 3-5+ year horizon)
conviction: 3
---

# Customer Analysis — Circle Internet Group (CRCL)

## Executive View
Circle's "customer" is a two-layer construct that most investors conflate. Layer 1 is the **distribution partners** that earn the revenue (Coinbase, Binance, Bybit, ecosystem apps) — and this is dangerously concentrated, with Coinbase alone consuming ~$908M of $1.01B FY24 distribution costs and ~22% of USDC float on platform. Layer 2 is the **end-user demand** for USDC float (DeFi protocols, exchanges, treasuries, payment apps, retail) — which is broad, growing fast (USDC in circulation +72% YoY to $75.3B in Q4 2025), but remains heavily skewed toward exchange settlement and DeFi collateral rather than the durable "payments rail" narrative being priced in. The bull case requires Circle Payments Network (CPN) and direct enterprise integrations to materially reduce dependence on the Coinbase rev-share before its 2026 contract auto-renewal cycle. The bear case is that Circle is currently a Coinbase financial subsidiary that pays away more than half of every dollar earned to a single partner whose negotiating leverage is structurally rising.

## Customer Concentration

### A. Distribution-partner concentration (the revenue side)
| Metric | Latest | YoY change | Source |
|---|---|---|---|
| Top distribution partner (Coinbase) % of distribution costs | ~90% in FY24 ($908M / $1.01B) | Up | insights4vc |
| Top distribution partner share of USDC float | Coinbase ~22% (Q1 2025), up from 12% (2023), 5% (2022) | Rising sharply | Coin Metrics |
| Named partners receiving rev share | Coinbase (50/100% split), Binance ($60.25M upfront + monthly SOFR-linked %), Bybit (terms undisclosed) | Bybit added 2025 | CoinDesk, Decrypt |
| FY25 total Distribution, Transaction & Other Costs | $1.66B+ implied (Q4 alone $461M, +52% YoY) | Up significantly | Circle Q4 2025 release |

### B. End-user float concentration (the demand side)
Top USDC-holding entities as of 2026 (Arkham):
| Rank | Entity | Type | Holdings | % Supply |
|---|---|---|---|---|
| 1 | Binance | CEX | $10.2B | 13.3% |
| 2 | Coinbase | CEX | $8.8B | 11.5% |
| 3 | Sky (MakerDAO) | DeFi | $3.7B | 4.9% |
| 4 | Hyperliquid | DEX | $3.3B | 4.3% |
| 5 | Circle (treasury) | Issuer | $2.6B | 3.4% |
| 6 | OKX | CEX | $1.1B | 1.4% |
| 7 | Polygon (bridge) | L2 | $1.1B | 1.4% |
| 8 | Bybit | CEX | $0.87B | 1.1% |

**Top-5 entity float concentration ≈ 37% of supply. Top-10 ≈ 43%.** Centralized exchanges alone control ~28% of float. This is materially more concentrated than a typical operating company's customer book — and crucially, the float concentration **mirrors the revenue-share recipient list**, which is the central structural problem.

## End-Market Exposure
USDC demand decomposes into roughly the following segments. Circle does NOT publicly disclose a clean breakdown, so percentages below are analyst estimates triangulated from on-chain data, transcripts, and third-party trackers — **all marked (est.)**:

| Segment | % of float (est.) | % of velocity (est.) | Cycle position | Structural | Macro sensitivity |
|---|---|---|---|---|---|
| CEX trading collateral | 30-35% | 40-50% | Mid-cycle (crypto bull) | Mature, plateauing share | Very high (crypto beta) |
| DeFi collateral / lending / DEX LP | 25-30% | 25-30% | Mid-cycle | Structural growth | High (rates + crypto) |
| Cross-border payments / remittances | 8-12% | 10-15% | Early-cycle | Structural growth (CAGR ~35%) | Medium |
| Institutional treasury / corporate | 5-10% | 5-10% | Very early | Structural growth | Medium-low |
| Tokenized RWA / Treasury settlement | 3-5% | <5% | Very early | Structural growth | Medium |
| Retail apps (Robinhood, Nubank, Meta creator pay, Intuit) | <5% | <5% | Very early | Optionality | Low |
| Bridges / cross-chain / market-maker float | 5-10% | 10-15% | Mid-cycle | Tied to chain proliferation | High |

**Weighted demand picture:** The reported "Circle is becoming a payments rail" narrative is directionally true but quantitatively early. Today, the **majority of USDC float and velocity still serves crypto-trading and DeFi infrastructure** — both of which are correlated to crypto market cycles and rate levels. The non-crypto-cyclical buckets (payments, treasury, RWA, retail apps) are growing fast but collectively are likely <25% of float. CPN doing $5.7B annualized in Q4 2025 (now $3.4B per latest Circle disclosure with mix shifts) is a ~0.05% sliver of USDC's $11.9T annual velocity. The 3-5 year compounder thesis depends entirely on these non-crypto buckets compounding from "rounding error" to "majority share." That is achievable but is not yet evidence; it is hypothesis.

## Contract Structure & Switching

**Coinbase Master Services Agreement (the dominant "contract"):**
- 100% of reserve income on USDC held on Coinbase platform
- 50% of reserve income on USDC held everywhere else (a structurally bizarre split given Coinbase doesn't issue or back the asset)
- Auto-renews on 3-year cycle; next renewal due late 2026
- Effectively makes Circle a 50/50 economic JV with Coinbase on the entire USDC franchise — Coinbase's economics improve as its platform share rises (and it's been rising)

**Binance agreement (Dec 2024):**
- $60.25M upfront cash payment from Circle
- Monthly incentive: mid-to-high-double-digit % of a fixed SOFR-linked rate, scaled to USDC balances on Binance
- Far more favorable to Circle than the Coinbase deal — Circle keeps majority of reserve income

**Bybit agreement (2025):** Confirmed rev-share, terms not public.

**Switching dynamics on the demand side:**
- For a CEX/DeFi protocol: USDC is sticky once integrated as quote currency, collateral standard, or pool LP — moving to USDT or competing stablecoin requires re-quoting books, re-tooling smart contracts, re-onboarding LPs. Switching cost is real but not insurmountable (Hyperliquid has gone the other direction; Tether retains 70%+ in Asia despite USDC's compliance edge).
- For a payments fintech (Stripe, Visa, Nubank, Robinhood, Intuit): integration is multi-quarter compliance + engineering work. Once embedded as the settlement asset, switching means re-architecting the rail. Higher stickiness.
- **Forward visibility / "RPO equivalent"**: Circle has none in the SaaS sense. The closest proxy is USDC float trajectory. Q4 2025 +72% YoY in float and +247% YoY in on-chain volume is a strong "burn rate" signal — the float is being USED, not just parked.

## Demand Quality

**Pull-through (sustainable, structural):**
- DeFi protocol integration as default collateral (Aave, Morpho, Spark, Compound)
- CCTP V2 cross-chain — USDC is now natively issued on 28 chains, making it the de facto multi-chain dollar (real switching cost for any app that builds on this primitive)
- MiCA compliance in EU — USDC is the only top-10 stablecoin compliant (12% EU market share, up from 10%)
- Visa USDC settlement (Dec 2025), Stripe x402 agentic-AI payments, Meta creator payments, Intuit (TurboTax/QuickBooks/Credit Karma), Nubank Brazil rollout

**Channel fill / re-mint churn (low-quality):**
- $257.5B minted in FY25 vs. $226.1B redeemed = $31.4B net new — but the redemption ratio (~88%) shows USDC behaves as **transactional float, not store-of-value**. Most users hold for days/weeks, not quarters. Reserve revenue is a function of **average float**, not gross issuance, so high mint/burn is fine — but means demand is rate-sensitive and arbitrage-driven, not religious.

**Pre-buy / yield-seeking (reverses if rates fall):**
- Reserve income at ~88% of revenue means USDC's economic engine = float × T-bill yield. If Fed cuts 200bps over the holding period, even with float doubling, revenue could be flat. This is the central rate-cyclicality risk on the demand quality of the float — but that belongs more to financial-analyst's lane. Customer-side observation: the marginal USDC holder is not buying it for yield (they get none directly), so demand is more rate-resilient than it looks; what's rate-sensitive is the unit economics, not the demand itself.

**Critical demand-quality fact:** Q4 2025 on-chain volume of $11.9T was +247% YoY while float was only +72%. Velocity is dramatically outpacing float growth. This is a **good signal** — it means USDC is being used more intensively per dollar issued, which is consistent with payment-rail adoption (high-velocity) versus passive parking (low-velocity). This is the single most underappreciated data point in the Circle bull case.

## Distribution-Partner Economics (Coinbase Specifically)
Coinbase is the customer concentration story, period. Walk the math:
- FY24: Coinbase received $908M of $1.01B in Circle's distribution costs = 90% of partner spend
- Coinbase's USDC platform share has roughly **doubled every ~18 months** (5% → 12% → 22%)
- Coinbase keeps 100% of reserve income on USDC on its platform → as that share rises, Circle's economics on that slice fall to zero
- The 50% rev-share on off-platform USDC means even if Circle wins a Stripe/Visa integration, Coinbase still takes half the reserve income from that float
- The 2026 auto-renewal is the **single most important customer-side event** for the long-term thesis. If Circle re-cuts the deal on more favorable terms (or lets it expire), this is a step-change in unit economics. If Coinbase keeps or expands its rights, Circle is structurally capped.

For a 3-5+ year compounder lens: this Coinbase relationship is the equivalent of **a SaaS company where its largest channel partner takes 50% of net revenue and that partner also operates a competing distribution channel** (Coinbase's own Base L2, USDC-centric ecosystem investments). Long-term holders need a credible path to Coinbase share **stabilizing or declining** as a % of float.

## Geographic Mix
- **United States:** core market; benefit from GENIUS Act regulatory clarity (2025); USDC is the default compliant US dollar token
- **EU/EEA:** USDC has MiCA compliance edge over Tether — but only translated to ~12% market share (vs. 10% pre-MiCA). Euro-denominated stablecoins are <0.1% of global stablecoin cap, so EURC is structurally niche
- **LATAM (Brazil, Argentina, Mexico):** 42% of new crypto-payment integrations; USDC growing fastest here via Nubank partnership and remittance corridors. Highest structural opportunity
- **APAC:** Tether dominates (~99% of cross-border crypto-USD volume in Asia is USDT). USDC is the laggard; Binance partnership is the wedge. Slow grind, not a step-change
- **Africa/Middle East:** Early CPN corridor (Nigeria announced); negligible today

Geographic verdict: USDC's regulatory advantage is a real moat in US/EU but is **not yet translating to share gains in the largest non-US growth markets** (Asia, where Tether's incumbent network effects are entrenched). LATAM is the best near-term geographic story.

## Customer Acquisition Flywheel
Each new institutional integration **does compound**, but in a non-obvious way:
- Direct revenue from a single fintech partner is small (Nubank, Intuit, Robinhood are unlikely to each move >1-2% of float)
- The compounding mechanism is **chain-of-trust**: each major brand integration (Visa, Stripe, Meta, Intuit) lowers the perceived risk of the next integration and shifts the default stablecoin choice for the next fintech evaluating the space
- This is a **slow-burn winner-takes-most flywheel**, not a viral one. It looks like Visa's merchant-acquirer flywheel circa 1985-2005: each year the network is incrementally less escapable

The flywheel is real but takes years. For a 3-5+ year holder, it is the central thesis. For a 12-month trader, it is invisible.

## Bull Points
- **USDC float +72% YoY and on-chain volume +247% YoY (Q4 25)** — the demand quality is improving (velocity outpacing float), which is consistent with a real payments-rail thesis, not just crypto-cycle beta
- **CPN at $3.4-5.7B annualized and growing** with 55+ institutions — small now, but represents the highest-margin (no Coinbase rev-share) demand bucket and is compounding
- **Regulatory moat in US (GENIUS Act) and EU (MiCA)** is locking in USDC as the default compliant dollar; Tether is structurally precluded from these channels
- **Distribution flywheel widening**: Visa, Stripe, Meta, Intuit, Nubank, Robinhood integrations in 2025-2026 each represent a new non-crypto demand bucket that grows independent of trading cycles
- **End-user switching costs are real** for any DeFi protocol or payment fintech that has built on USDC primitives (CCTP V2 cross-chain native issuance on 28 chains)

## Bear Points
- **Customer concentration on Coinbase is extreme and worsening**: 22% of float on a single platform, ~90% of distribution costs going to a single counterparty, and that counterparty's share has tripled in three years
- **Coinbase contract auto-renews late 2026** — Coinbase's leverage at the negotiating table is materially higher than in 2023; the most likely outcome is terms get worse for Circle, not better
- **Float is highly entity-concentrated**: top 5 entities = 37% of supply, top 10 = 43%. This is real run risk: a single major exchange de-listing or delpegging fear triggers redemption cascade (March 2023 SVB episode is the reference scenario)
- **Tether dominates Asia (~99%) and globally (~70%+ market share)** — USDC's MiCA/GENIUS edge has not translated to global share gains; geographic ceiling exists
- **Demand is still mostly crypto-cyclical** — the durable payments/treasury/RWA buckets are <25% of float despite being 90% of the bull narrative; reality is well behind the story
- **Mint/redeem ratio (88%) reveals USDC is transactional, not store-of-value** — float is rate-sensitive in unit economics and trading-cycle-sensitive in volume

## Conviction (1-5)
**3 / 5** — Mixed. The customer-side picture is genuinely bifurcated: end-user demand is broad, growing, and structurally improving (velocity > float growth, real regulatory moat, real CPN traction). But the revenue-side customer concentration on Coinbase is so extreme that it materially caps the long-term economics regardless of how successful USDC becomes as an asset. For a 3-5+ year compounder lens, this is a "yes but" — Circle's customer base is durable and expanding, but the partner concentration is the single overriding factor. Conviction would move to 4 if Coinbase rev-share renegotiation goes Circle's way in late 2026, and to 2 if the Coinbase share of float crosses 30%.

## Key Risks to This Read
1. **Coinbase 2026 contract renewal** — terms could worsen materially; Coinbase has every incentive to extract more given its rising platform share
2. **Float concentration tail risk** — if Binance has a regulatory event or de-lists USDC, ~13% of supply could redeem in days
3. **CPN may be a vanity metric** — $3-5B annualized is meaningful PR but immaterial vs. $11.9T total volume; need to see real take rates and stickiness data
4. **Asia geographic ceiling** — Tether's network effects in APAC may be permanent; USDC's TAM may be structurally smaller than the bull case assumes
5. **Demand-quality misread risk** — the velocity-outpacing-float datapoint could partly reflect MEV/wash-trading on DEXes, not "real" payment activity. Need transaction decomposition Circle does not yet provide
6. **Retail fintech integrations may not monetize** — Robinhood, Nubank, Meta integrations look great in press releases but their float contribution remains <5% combined

## Sources
- Circle Q4 2025 earnings press release (March 2026)
- Circle Q3 2025 earnings press release
- Circle 10-K for FY 2025 (filed March 9, 2026)
- Circle Payments Network whitepaper and product launch (April 2026)
- insights4vc / "Inside Circle's Stablecoin Economics"
- Coin Metrics State of the Network Issue 317 (Circle valuation deep-dive)
- Decrypt: "Coinbase Takes 50% Share of Circle's Residual USDC Reserve Revenue"
- CoinDesk: "Circle Has USDC Revenue Sharing Deal With ByBit" (July 2025)
- Arkham Intelligence: "Who Owns USDC — Top USDC Holders 2026"
- Stable100: top wallets holding USDT and USDC
- Kaiko Research: "MiCA is Reshaping EUR Stablecoin Markets"
- East Asia Forum: "Market stacked against East Asia's stablecoins" (Jan 2026)
- DefiLlama stablecoin supply data
- Bessemer Venture Partners: "Stablecoins: from DeFi primitive to global financial infrastructure"
- Circle press releases on Robinhood, Nubank, Visa, Intuit, Stripe partnerships
- Coingeek: "Circle's 2025 ended on a high"