Visa lost its stablecoin engine. Then it placed a desperate bet on a new partner. The RFP is not just a procurement document. It is a confession.
For 18 months, Visa had a quiet, powerful relationship with BVNK. It invested in May 2025 at a ~$750M valuation. It integrated direct stablecoin settlement via Visa Direct. The architecture was clean: Visa controlled the front-end rails, BVNK handled the back-end OTC and multi-stablecoin conversion. It was a functional, symbiotic stack.
Then Mastercard intervened. On March 17, 2026, it announced the acquisition of BVNK. By August 3, the deal closed at up to $1.8B. In nine months, BVNK’s value multiplied by 2.4x. Not because of improved fundamentals. Because Mastercard paid a strategic premium to break Visa’s infrastructure.
Visa had 13 days. On August 5, it integrated Zero Hash as a stopgap to keep stablecoin settlement alive across 195 countries and 18 billion endpoints. Then on August 18, it released the RFP that would define its next decade.
The architecture of trust is built, not inherited.
The RFP Reveals the Real Problem: It’s Not Just a Vendor Search
The RFP asks for four critical licenses: US, Canada, UK, Singapore. It demands the ability to convert and support multiple stablecoins. It requires the capacity to handle OUSD settlement loads. At first glance, this reads like a standard compliance checklist.
Look closer. The licenses are a regulatory moat, but they also signal a deeper structural demand: Visa is not looking for a supplier. It is looking for a risk-sharing partner. The OUSD Alliance, which includes 140+ companies from BlackRock to Amex, requires a stablecoin backend that can manage multi-party coordination, not just multi-currency conversion. The RFP’s multi-stablecoin requirement is not a feature request. It is a governance requirement.
This is the core insight: Visa is not rebuilding a backend. It is reconstructing a trust network. The previous architecture depended on a single node (BVNK). Mastercard severed that node. Now Visa must find a partner that can serve as a neutral, regulated, multi-stablecoin settlement layer for a coalition of competitors.
The Hidden Details: Three Dimensions of the RFP That Most Analysts Miss
Dimension 1: The Multi-Stablecoin Trap. The RFP demands conversion across multiple stablecoins. This sounds simple. It is not. Real-time, low-slippage conversion between USDC, USDT, and OUSD at scale requires deep liquidity pools and sophisticated risk management. Most OTC desks can handle one pair. Few can handle an open matrix. The RFP is a filter that eliminates 90% of naive candidates.
Dimension 2: The Zero Hash Interim. Zero Hash is a compliance API layer. It is not a full OTC settlement backend. The integration on August 5 was a bandage, not a fix. The 13-day gap between Zero Hash integration and the RFP release suggests Visa had a fallback plan, but it was not a long-term strategy. Zero Hash likely lacks the multi-stablecoin liquidity depth and multi-jurisdiction license stack that the RFP demands. It is a temporary bridge, not a permanent foundation.
Dimension 3: The OUSD Solana Timeline. OUSD plans to launch on Solana in H2 2026. Zero fees on minting and redemption. Revenue flows to distribution partners. This model is structurally fragile. Zero fees are sustainable only if reserve asset yields remain high. In a declining interest rate environment, the economics break. The RFP candidate must understand this fragility and provide contingency mechanisms. Most candidates will not have the institutional experience to manage this.
The Contrarian Angle: Mastercard’s Victory May Be Pyrrhic
The conventional narrative is that Mastercard won. It paid $1.8B for BVNK, vertically integrated stablecoin settlement into Mastercard Move, and now operates 24/7 on-chain settlement. It looks like a decisive strategic victory.
But there is a hidden cost. Mastercard now owns the backend. That means it also owns the liability. If BVNK’s infrastructure fails, if a stablecoin de-pegs, if a regulator challenges the licensing structure, Mastercard bears the full weight. Vertical integration provides control, but it also removes the buffer of neutrality.
Visa’s alliance model, by contrast, distributes risk across the OUSD coalition. Visa is not the backend operator. It is the settlement facilitator. This is a more defensible position in a regulatory environment that increasingly targets infrastructure providers. Mastercard consolidated its infrastructure. Visa diversified its trust network.
The question is not which model is faster. It is which model is more resilient to the next regulatory shock. Based on my experience auditing 12 ICO whitepapers in 2017 and surviving the 2022 bear market by stress-testing Layer 2 protocols, I know that resilience outlasts speed in crypto infrastructure.
The Takeaway: Two Futures, One Winner
The RFP will determine the next phase of stablecoin payments. If Visa finds a partner that can meet its multi-license, multi-stablecoin, multi-party requirements, the alliance model will prove more durable than Mastercard’s vertical stack. If the RFP fails, Visa’s stablecoin momentum stalls, and Mastercard consolidates its lead.
Watch the OUSD Solana launch. Watch the RFP candidate announcement. The architecture of the next payment system is being built right now. The question is: who will be the architect?