Why Mercury Is the Neobank That Ate SVB — And Why Its $3B Valuation Is a Rate-Cut Bet in Disguise

On 8 June 2026, the technology-business talk show TBPN relayed a Bloomberg scoop from Kate Clark: Sequoia Capital is leading a primary investment in Mercury, the startup-focused banking platform, at a valuation north of $3 billion — roughly double the company's last private mark of $1.6 billion in 2021. The reported financials are striking. Mercury is doing $500 million in annualized revenue and roughly $200 million of EBITDA, a margin north of 40%, on a customer base dominated by venture-backed companies, e-commerce merchants and, increasingly, the long tail of small firms that lost patience with their bank during the 2023 regional-banking crisis.

The story of how Mercury got here is the story of how the Silicon Valley Bank collapse was privatised — and how a neobank, by deliberately refusing to take a bank charter, has quietly become a piece of systemically important startup infrastructure that no regulator currently oversees as a bank.

A neobank by design, not by accident

Mercury is not a bank. It has never wanted to be one. The strategic logic, as laid out on TBPN, is straightforward and somewhat contrarian: a bank charter forces an institution to diversify its customer base, build out lending infrastructure, and operate under FDIC-supervised capital and risk regimes that would erode the high-margin deposit-gathering business Mercury has spent five years perfecting. By partnering with a network of FDIC-insured banks — and offering customers an aggregate deposit coverage that no single bank can match — Mercury keeps the regulatory ceiling low and the unit economics high.

"On $500M of annualized revenue they put up $200M of EBITDA… the reason they're not getting full credit is because the vast majority of their earnings are from interest yield… if rates drop over the next few years a lot of that profit will just dry up."

That arithmetic — fat margin, yield-dependent — is the entire investment thesis, and it is the entire risk.

The SVB harvest

Mercury's acceleration is unintelligible without March 2023. When Silicon Valley Bank failed, the Federal Reserve and Treasury stepped in to backstop uninsured depositors — but the lesson the startup ecosystem actually absorbed was not "the system worked." It was "find a banking partner that is not a single regional bank with concentrated deposit risk." Mercury was one of the few platforms that could absorb fleeing SVB customers without disruption, and it could advertise expanded FDIC coverage precisely because it routed deposits through multiple partner banks. The TBPN hosts described the dynamic bluntly: a portfolio company in their own network had been offering 4% yield on deposits through a single bank partner and lost customers overnight when SVB collapsed. Mercury, structurally, was built to never be that story.

"Once you have a [bank] charter you're getting highly highly regulated and they will force you to diversify your customer base… Mercury was able to hit a few of these waves — the e-com wave, the general growth in startup investment, and then the post-SVB being there to catch the companies that needed a new bank partner."

This is a textbook platform-economics story dressed in a banking costume. Mercury did not invent yield on deposits. It invented the ability for a Series A SaaS startup to get 4% on its operating cash in five minutes, with a dashboard, and to feel that the money is somehow safer than it would be at Wells Fargo.

The $3 billion is a Fed bet

Here is the part the Bloomberg headline does not capture. Mercury cannot command a SaaS multiple on $500 million of annualized revenue and $200 million of EBITDA, because the market knows that the EBITDA is, mechanically, a function of the federal funds rate. The TBPN analysts put it crisply: a 40%-plus EBITDA margin at a software company would justify a 20-times multiple; at a deposit-gathering platform whose earnings track the front end of the yield curve, the multiple compresses. Sequoia is buying a real business with real customers and a real moat — and is also implicitly underwriting a path in which the Fed does not cut as aggressively as the forward curve currently implies. If the Fed holds rates higher for longer, Mercury's earnings power holds. If the Fed cuts into the 3% range, the $200 million of EBITDA starts to look more like $150 million, then $120 million, and the valuation math that justifies $3 billion today starts to look thin.

This is not a flaw unique to Mercury. It is the defining risk of the entire B2B-fintech cohort — Brex, Ramp, Mercury, the deposit-flotation layer of the neobank stack — that built scale during 2022–2024 on the back of the most aggressive rate-hiking cycle in four decades. But Mercury is the cleanest expression of the bet, because it has refused every opportunity to diversify away from it.

The structural frame: payroll, deposits, and the unglamorous centre of the economy

Zoom out and the story gets stranger. Three of the top ten most valuable Y Combinator companies — Rippling, Gusto, Deel — are payroll providers. ADP alone does roughly $20 billion in trailing-twelve-month revenue and around $6 billion of EBITDA, a reminder that "follow the money" businesses built on a tiny sliver of total wages produce enormous enterprise value. Andreessen Horowitz, which seeded both Mercury and Deel, is openly building a portfolio that monetises the boring connective tissue of how companies move cash and pay people.

Mercury sits one layer beneath that. It is the bank account those payroll companies, those e-commerce merchants, those AI startups — every one of them a customer of a neobank that is itself a customer of partner banks that are themselves customers of the Fed — ultimately settle into. The interesting question is not whether Mercury is a good business. At $500 million of ARR with the customer cohort it has, it plainly is. The question is what it is worth in a world where the Fed has already won its inflation fight and is now preparing to ease.

Stakes: a chartered question waiting for a crisis

The structural concern is not that Mercury fails. Mercury is well-capitalised, profitable, and operationally mature. The concern is what happens the next time a credit shock hits the venture ecosystem and a platform that holds the operating cash of tens of thousands of startups is not, itself, a bank. The 2023 SVB episode was resolved because the uninsured depositor backstop held. A 2027 episode in which the backstop does not hold, and in which the deposit-taker is a Series D fintech with a partner-bank model, is a different conversation. Mercury has effectively privatised the upside of the deposit franchise and socialised the contingency. That is a great business to own. It is also the kind of arrangement that ends in a regulatory letter after the next crisis, not before it.

Sequoia's $3 billion bet, in other words, is a rate bet, a distribution bet, and a quiet wager that the regulatory perimeter around deposit aggregators does not move before the next liquidity event. Mercury has built the neobank that ate SVB. The question is who eats Mercury if the curve cuts faster than its model assumes.

Wire provenance

This editorial synthesis draws on the following public wire/social posts:

• https://www.youtube.com/watch?v=o2-OGvpNjQA