For more than a decade, fintech’s thesis was simple: build what banks wouldn’t – instant credit at checkout, mobile-first money management, frictionless payments, without carrying the weight of being a regulated institution. They thrived by renting bank infrastructure (deposits, charters, compliance) while owning the customer touchpoints. And that served them well, until it didn’t.What changed is that “rent-a-bank” stopped feeling like a shortcut and started feeling like a ceiling. As fintechs scaled, the partner model increasingly meant:higher structural funding costs vs. insured deposits, shared economics on the most valuable layers (deposits, lending margin, network access), and operational fragility when a sponsor bank’s risk appetite, or a regulator’s interpretation, shifted. For countries like the US, these pressures are amplified by a bank-first system that’s also more concentrated. The number of US banks fell from 8,500 (2008) to ~4,500 (2025). This raises the stakes around who gets to enter the system and on what terms. Trust still tilts toward incumbents: 68% of the general population trust banks vs. 53% for fintechs (and 43% for crypto/digital assets). And so do unit economics: chartered fintechs average 14–16% ROE vs. 9–10% for non-banks, largely because insured deposits are cheaper than wholesale or partner-bank funding (often adding ~1.5–3% to […]