⚠ A Flywheel That Also Concentrates the RiskModerate threat
StoneCo (STNE) — threat to the moat
When a merchant fails StoneCo loses the volume, the deposit and the loan at once, because they were never separate exposures.
The attraction of the integrated model is that each product makes the next one better: payments generate the data, the data underwrites the credit, the credit funds the merchant, and the merchant's growth generates more payments. It compounds in the good direction.
It also compounds in the other one. When a merchant fails, StoneCo does not lose one relationship. It loses the payment volume, the deposit balance, the loan, and the acquisition cost spent to win all three — and it loses them simultaneously, because they were never separate exposures. The company's own filing warns that the concentration of its clients by geography and economic sector may increase its risk, and it lists churn from business closures as a standing feature of a base composed of small merchants1.
Brazilian small businesses are a genuinely difficult customer set to be levered to. They are sensitive to the policy rate through their borrowing, to unemployment through their sales, and to consumer credit conditions through their customers — three variables that move together and are moving now.
Nothing here says the flywheel is a bad design. It is the right design, and it is why the moat page rates it as it does. The point is narrower: an integrated relationship converts several small independent exposures into one large correlated one, and correlation is what turns a difficult year into a bad one.
Watch the loss rate against the Brazilian small-business default rate. A flywheel is supposed to underwrite better than the market, not merely to be more exposed to it.
- ReportedStoneCo's filing warns that the concentration of its clients by geography and economic sector may increase its risk, and that it experiences churn from business closures, account transfers and declines in same-store sales.StoneCo Ltd. Form 20-F for FY2025 (CIK 1745431) — active payment clients of 4,803.5 thousand at 31 December 2025, against 4,172.7 thousand in 2024 and 3,522.1 thousand in 2023; TPV of R$560.9 billion, against R$516.2 billion and R$438.3 billion; revenue of R$14,153.8 million and adjusted net income from continuing operations of R$2,477.2 million; more than 3.6 million banking active clients, the majority of whom are also payment clients; retail deposits of R$11,091.0 million against R$8,704.8 million and R$6,119.5 million; a credit portfolio of R$2,836 million with expected credit losses of R$389.7 million, against R$1,207.6 million and R$144.5 million a year earlier; clients divided into MSMBs (micro-merchants and SMBs) and Key Accounts, 'comprised of platform services and sub-acquirers'; StoneCo became in 2017 the first non-banking entity authorised by the Central Bank to operate as an Acquirer through a payments-institution licence, and is among the six largest players by total card volume per ABECS; distribution through proprietary and franchised hubs sold on 'service differentiation as the main driver', digital channels, and more than 500 Strategic Partners at December 2025; per the Central Bank, Pix's share of the total number of transactions rose from 1% in Q4 2020 to 52% in H1 2025 and its share of monetary volume from 1% to more than 26%; the filing warns that 'the concentration of our clients by geography and economic sector may increase our risk' and that the company experiences churn from business closures and account transfers; interest rates directly affect both revenue generation and cost of funds, most third-party funding being linked to the Brazilian interbank rate; StoneCo's own analysis notes that US MSMB take rates have been stable over five years despite penetration around 120% of consumption, and finds no indication of saturation-driven price reductions in Brazilian cities with low cash usage — FY2025 · publ. 2026 · source ↗