Most fintech PMs look at Mercury’s Treasury product and see a yield offering. That’s the wrong unit of analysis. Mercury Treasury is a study in how to stack three regulatory regimes — banking, investment advice, and broker-dealer custody — inside a single UX without collapsing them into legal mush.
If you’re scoping any cash-management, yield, lending, or investment product in fintech, the architecture Mercury ships is the architecture you should steal. Let’s break it down.
What the product does
Mercury Treasury moves idle cash from a Mercury banking customer’s checking account into one of two low-risk vehicles: the J.P. Morgan U.S. Treasury Plus Money Market Fund or the Morgan Stanley Ultra-Short Income Portfolio. Yields run 3.02% to 3.65% depending on balance. Minimum $250,000 across all Mercury accounts. Funds move into brokerage accounts held in the customer’s name at Apex Clearing Corp, a third-party broker-dealer.
Money can return to the Mercury checking account same-day from the money-market fund (if initiated before 3pm ET) or within 1–4 business days from the ultra-short bond fund. Fees: 0.15–0.6% annual on Treasury balances. No transaction fees.
What’s strong
The three-entity architecture is legally clean. Mercury (fintech) owns the UX. Mercury Advisory LLC (SEC-registered RIA) owns the investment advice. Apex Clearing (SEC + FINRA broker-dealer) owns custody. Each entity holds exactly one license. Customer funds are separated from Mercury’s banking operations and survive Mercury’s hypothetical bankruptcy. This is how you build fintech that can be audited without excuses.
The yield table is transparent and tiered. Six explicit tiers from $250k to $20M+, separate rates for each of the two products. A customer can compute their exact APY before moving money. Most competing treasury products bury this behind “contact us” or blended rates. Mercury published it.
The same-day settlement on the money-market fund is genuine. Most fintech “same-day” promises dissolve under the rail’s real cycle. Mercury’s cutoff is explicit (3pm ET), the fund itself is daily-liquid by design, and the user sees the constraint before picking the product.
What’s weak
The eligibility list is a silent feature-killer. Customers must be a U.S. entity, physically located in one of 11 listed countries. Sole-proprietor LLCs, nonprofits, foreign financial institutions, certain investment advisers, and securities brokers are excluded. For a PM building signup flows, the volume of customer types filtered out at the “Open Treasury” click is higher than Mercury’s marketing implies — and the exclusions are not surfaced until the gating screen.
The 1–4 business-day settlement on the Ultra-Short Portfolio is a ticking clock. The fund yields more (3.65% vs 3.47% at the top tier) but locks money up longer. For working-capital use cases (payroll, supplier payments), the delta is not worth the yield bump — yet the UI places both products side-by-side as parallel choices.
SIPC coverage is framed as insurance, but it isn’t the same thing as FDIC. SIPC covers brokerage failure, not market loss. A customer reading the page quickly could conclude their Treasury balance is “insured up to $500k” in the same sense their checking is “insured up to $5M via FDIC.” The disclosures are legally complete but perceptually misleading for customers who don’t already know the difference.
The compliance angle — which no other teardown will add
Mercury Treasury reads like someone drew a Money Flow Map first and built the product around the diagram.
The critical move: Mercury did not try to offer Treasury “as Mercury.” They stood up Mercury Advisory LLC as a separate legal entity with its own SEC registration, and they partnered with Apex for custody rather than trying to hold funds themselves. A newer fintech’s instinct would have been to market Treasury as a native feature — “Mercury Yield” — and quietly license a partner. Mercury did the opposite: surfaced the three-entity reality into the disclosures, and let Apex’s 40-year broker-dealer history do the trust work.
For a fintech PM scoping a similar product, the implication is direct: the yield product is not your product. The brokerage partner is your product. The license architecture is your product. The Mercury UX is the thinnest possible layer on top.
The tension this creates is that the UX looks unified but the legal reality is fragmented. A customer who doesn’t read the disclosures won’t realize they’ve switched from FDIC to SIPC territory the moment they move money. Mercury accepts this tension in exchange for a clean legal posture. Newer fintechs often get this backwards — they smooth the UX, but when a regulator pulls the thread, the entity structure unravels.
What a PM should steal
The three-entity pattern. If you’re scoping any cash-management, yield, lending, or investment product, map your three entities (or two, or four) before you draw a screen. The Money Flow Map should have at least two licenses on it. If you only have one, you are either over-licensing into one entity or building something your license doesn’t cover.
Tiered pricing in public. Most competitors hide pricing behind sales calls. Mercury lists it. The revenue cost of transparency is near-zero (sophisticated buyers calculate anyway). The trust payoff is large — and it serves as a continuous internal pressure test on whether the economics still make sense.
What a PM should avoid
The “parallel product” anti-pattern. Placing Money Market and Ultra-Short Bond as equivalent choices on the same screen forces the customer to understand T-bills, commercial paper, settlement cycles, and yield/liquidity tradeoffs before they can click. That’s a cognitive wall. The better pattern is a recommendation engine: ask two questions (“When might you need this money back?” and “What’s your risk tolerance?”), then pick one. Offer “switch to the other” as a post-click option.
What to do Monday
Pick one product in your fintech that holds, moves, or earns money. Draw the Money Flow Map. If you can’t identify two separate legal entities handling two separate regulatory regimes, either (a) you’re over-licensed in one entity — a risk — or (b) you’re under-licensed for what the UX promises — a bigger risk. Fix the map before the next roadmap review.
If you’re designing a yield, treasury, or investment-linked product right now and want a second pair of eyes on the entity architecture before engineering starts, that’s what my 2-week MVP scoping sprints solve. Book a strategy call: cal.com/barakazuga/strategy-call.
— Barak
The Regulated MVP is a weekly tactical brief for fintech and AI-in-finance product leaders. Every Tuesday.