Dave CEO, Jason Wilk: The Best Performing Fund Would Only Back YC Founders on Their Second Time
Summary
Wilk thinks richer founders often make better founders because a first exit buys confidence and downside protection to “swing for the fences” on company two. He argues that blindly writing uncapped notes into successful exited YC founders’ next companies would probably produce “one of the best VC funds on the planet.” His own first exit let him attack banking rather than protect a niche, cash-generating business.
Capital abundance can destroy that advantage when copycat founders accumulate preference stacks without “a real bone to pick” with the industry. Dave has considered acquisitions whose targets raised hundreds of millions in preferred capital, making modest exits impossible until the cash runs out. Wilk says founders resist those offers; venture investors, focused on 10–20x winners, generally do not care about merely recovering capital.
Wilk still prefers the SPAC’s fixed valuation and predictable PIPE capital to an IPO’s uncertain price and nine-to-12-month process; his regret is going public six-to-12 months too late. The host describes Dave’s January 2022 listing at a $4 billion market cap, while Wilk later refers to a $5 billion valuation before the collapse. Fintech, SPACs, and unprofitable growth all became toxic by April; every PIPE investor exited before lockup expiration, leaving Dave a “sitting duck” at a $50 million market cap. He thinks one high-quality listing could rehabilitate the structure.
Dave survived its 98% drawdown by treating the inaccessible IPO value as “never real” and organizing around “patience and performance.” Its mission kept departures low, while out-of-the-money performance stock units at stock-price milestones from $1 through $100 ultimately “minted more millionaires” than the IPO. The recovery reached roughly $1.13 billion at recording, though Wilk says, “I’m not feeling like we’re through the woods yet.”
Cash-flow AI transformed Dave’s credit economics: the average advance rose from roughly $50 at launch to $180 by year-end 2024 while losses fell from above 10% to 1.2%, versus an industry level Wilk puts above 5%. About 12 million connected accounts and nearly one billion transactions feed a model whose five-to-10-day advances mature quickly, allowing it to learn every few weeks. On $1.6 billion of Q4 originations, every 10 basis points of loss improvement matters, helping gross margin rise from the mid-40s in 2022 to 72%.
The broader turnaround came from operating leverage, AI, and contract cleanup—not layoffs: Dave had about 300 employees at its IPO and still had 300 afterward. AI handles routine support at a fraction of the roughly $2–$3 cost of an agent contact while the domestic escalation team stayed the same size. Monthly paying members reached the stated 2.1 million profitability threshold in Q4 2023, producing $10 million of EBITDA; at 2.5 million in Q4 2024, profitability reached $33 million, with 2025 guidance of $110–$120 million.
Dave’s distribution thesis is to deliver credit in five minutes and let the relationship grow from “snack” to “meal,” rather than paying heavily to make a new customer switch primary banks immediately. That produces a stated $16 CAC and 30% word-of-mouth acquisition, versus the capital-intensive direct-deposit strategy Wilk associates with Chime. He sees longer-duration credit—potentially $500 repaid over six paychecks—as the next use of Dave’s “CashAI,” not a race to subsidize free BNPL.
Wilk prefers competition to price regulation and says Trump is better for business because companies face less risk of tripping “some government wire.” He argues a 10% credit-card APR cap could shrink approvals and push borrowers toward payday loans, while eliminating overdraft fees could reappear as account fees or denied liquidity. As a public CEO, he is also skeptical of short sellers whose reports amplify assumptions for profit, arguing that a long-only market “wouldn’t look terribly different.”
Deep dive
1. Prior liquidity changes the size of a founder’s ambition
Wilk’s qualified answer to whether richer founders make better founders is “I’d say yes”—not universally, but often enough that an uncapped bet on every successfully exited YC founder’s second company could be an extraordinary fund.
His cleanest specimen is Eric Glyman: after selling Paribus to Capital One, Glyman mentioned another fintech idea and founded Ramp roughly a month later. The mechanism is financial safety plus earned confidence, not wealth alone.
Wilk’s first company fought for a $300,000 seed round in 2009–10; Mark Cuban capped his salary at $30,000 until profitability, and the company never raised again. That constraint taught capital efficiency—and repeated overdrafts on that salary supplied the personal grievance behind Dave.
The second time, Wilk had money in his pocket and no longer needed to protect a niche business in which roughly 40% of each cash dollar was his. He could accept failure and attack banking: “I had a real bone to pick with a major industry.”
2. Preference stacks can turn abundant funding into paralysis
Wilk’s critique of the 2021–22 cohort is not simply overvaluation: too many copycats raised heavily without “real skin in the game” or enough passion to endure the years required to build a company.
Large preference stacks then remove strategic optionality. Dave has considered buying small companies that raised a couple hundred million dollars; their eventual outcomes become impossible if no purchase price can satisfy that capital structure.
Harry’s pushback—worth keeping—is that eventually someone offers $20 million against $200 million raised and expects gratitude. Wilk agrees that reckoning will come, but only after unusually large cash reserves finally burn down.
Venture investors seldom force a recovery sale because merely getting their money back does not matter in a hit-driven portfolio; attention flows toward 10–20x fund returners. Wilk says the founders are generally the ones rejecting sub-stack outcomes.
3. Public markets solve real problems, but Dave reached them too late
Going public erased Dave’s preferred equity and left it with no debt; Wilk also cites roughly $100 million of daily stock volume as meaningful liquidity for employees. Only rare companies with abundant secondary demand and obvious public comparables can comfortably stay private.
Consumer companies may gain something enterprise businesses cannot: passionate customers becoming shareholders. His analogy is Tesla, whose owner-investor cult following, he argues, helped push its valuation beyond what a private-market EBITDA multiple would support.
The SPAC’s appeal was certainty: a set valuation, known dilution, and a guaranteed amount of PIPE capital, versus an IPO whose proceeds and price emerge only after an arduous nine-to-12-month process. Low-quality listings—not the mechanism itself—made “SPAC” poisonous. Wilk thinks a high-quality company could reset that stigma.
Wilk does not regret the vehicle; he regrets waiting six-to-12 months. Dave listed in January 2022, the market broke in April, and all PIPE investors sold before lockup expiration despite promising, “Stock goes down, we’re going to buy more.” Without seasoned holders or analyst coverage, Dave became a “sitting duck” or “fallen angel” at a $50 million market cap.
4. “Patience and performance” carried the company through a 98% drawdown
Public-market pressure did impose a cost: longer-term products were sidelined while Dave improved its core offering and margins. New bets were finally scheduled for later in the year and into 2026, but Wilk says they might already have launched without the collapse.
A friend supplied the coping mechanism: because Dave was down 90% before Wilk’s lockup expired, “It was never real.” There had never been a moment when he could actually take $100 million off the table; only the path forward remained actionable.
Internally, the message became “patience and performance.” Wilk says the mission—reducing the overdraft burden on everyday Americans—kept departures remarkably low because IPO wealth was supposed to be a byproduct, not the reason to work there.
Dave issued deeply out-of-the-money performance units tied to stock prices of $1, $5, $20, $80, and $100. Wilk believes those awards ultimately “minted more millionaires” than the IPO; employees who left left substantial money on the table. His wife’s only seed investment, a $50,000 Dave check, was up about 100x.
5. The crypto detour clarified where Dave’s real edge lived
Raising Dave’s Series A required about 120 meetings despite a $5 CAC and strong growth. One healthcare-oriented fund ultimately invested $10 million at roughly a $40 million price; the next round valued Dave at $1 billion, and only $60 million of primary capital funded it before the public listing.
Wilk still believes in blockchain, stablecoins, and Bitcoin as digital stored value, but calls crypto a distraction for Dave. Its FTX partnership coincided with the stock’s all-time high and included a $100 million convertible note, yet the product never launched.
Wilk characterized the note as debt rather than an equity investment, so Dave owed the money back regardless of FTX’s collapse. The note was not due until 2026, but Dave paid $71 million early; Wilk estimated that holding it would have meant repaying roughly $109–$110 million. He viewed the transaction as accretive and as a confidence signal before Dave reported its first profitable quarter. The company ultimately sidelined initiatives to double down on AI and its core business.
6. Fast feedback lets cash-flow AI bend the loss curve
Dave underwrites from cash-flow data by connecting accounts through Plaid. Roughly 12 million connected accounts contribute six months of initial history plus ongoing data, giving the platform access to nearly one billion transactions.
The original rules engine examined paydays, positive-balance persistence, and similar variables. It produced losses above 10% while offering $75, with the average near $50; by year-end 2024, the average was $180 and losses were 1.2%.
AI can find patterns that rules miss—employer, shopping behavior, ATM location, or fraud clusters. Because advances usually last only five-to-10 days, the full portfolio matures quickly and the model can learn every few weeks; an installment lender might wait six-to-12 months to evaluate a model vintage.
That creates the unusual divergence Wilk emphasizes: more credit per customer alongside lower losses, rather than safer underwriting requiring smaller limits. With $1.6 billion of Q4 originations, each 10-basis-point improvement adds material margin; gross margin advanced from the mid-40s in 2022 to 72%.
7. AI turned a fixed platform into operating leverage
AI now resolves routine support questions in seconds and, Wilk says, produces better scores for a fraction of the cost. A human interaction costs roughly $2–$3 or more; an AI response costs “literally nothing” by his framing.
This was substitution through scale, not a domestic layoff story. Dave retained roughly 300 employees from IPO through the turnaround while its customer base doubled; the US escalation team stayed constant, with reduced dependence on outsourced support. Renegotiated processing, network, and infrastructure contracts supplied additional savings.
Wilk contrasts JPMorgan’s cited cost of about $300 annually to break even on a basic checking account with Dave’s roughly $40 cost to serve. Legacy costs force incumbents to recover expenses through cross-selling, minimum balances, or $35 overdraft charges; Dave can offer free checking and charge about $5 to access a $100 advance.
Management told investors the platform would become profitable at 2.1 million monthly paying members. That arrived in Q4 2023 with $10 million of EBITDA; 2.5 million members produced $33 million of profitability in Q4 2024, followed by 2025 guidance of $110–$120 million. Wilk’s philosophical shift was from “growth at all costs” to “profitability at all costs.”
8. Credit-first distribution turns a “snack” into the banking relationship
Dave promises up to $500 within five minutes: a customer connects an existing account, receives an underwriting decision, tries the debit card, and may move direct deposit later. Wilk says that speed-to-value produces a $16 CAC, with 30% of acquisition coming through word of mouth.
Chime’s contrasting goal is to become the primary bank immediately, a proposition Wilk considers expensive because customers dislike moving bills and direct deposit to an unfamiliar institution. Dave reached IPO on $60 million of primary capital by letting repeated “snacks” become the meal.
Scale compounds the underwriting moat: customers have used ExtraCash roughly 130 million times, continuously enriching repayment data while higher volume lowers network and servicing costs. The next step could be longer-duration credit—for example, $500 repaid across six paychecks—for travel, schoolbooks, or other discretionary purchases.
A global digital bank is increasingly feasible as Plaid reaches 14 countries and companies such as Bridge and Stripe support innovation around stablecoins that could reduce cross-border currency friction. Wilk says Revolut often succeeds where incumbent banks lack mobile apps, while Nubank’s customers are middle- to higher-income; Dave and Chime target poorly served US customers.
In the US, Wilk sees banking as adequate for many earning above $100,000, while the opening is among younger, lower-income customers that incumbents serve expensively. He characterizes the opportunity as roughly 50% of Americans earning below $100,000 and living paycheck to paycheck. His advice to Revolut is to identify the worst-served segment or accept CAC approaching $500.
9. Wilk favors competition over blunt regulation—and distrusts pedigree and shorts
Asked directly whether Trump is better for business than Biden, Wilk says yes because reduced regulation lets companies innovate without fearing they will “trip some government wire.” He calls the case against Dave—filed on Election Day after good-faith negotiations—government overreach and argues that 14,000 banks plus 50 neobanks give consumers ample exit options.
A 10% credit-card APR cap, in his reasoning, would not remove risk; it could shrink approvals and push excluded borrowers toward payday loans. Likewise, banning overdraft charges could prompt banks to increase monthly account fees or deny a liquidity lifeline: “It would suck to get stuck at the gas station just because a government regulator wanted to get a headline win.”
His founder-management regret is overhiring pedigreed C-suite executives because venture capital encourages it. Previous-company success is not cultural fit; rushed senior hires can disrupt a company both while present and when their visible departures unsettle everyone else.
Wilk declines to name a short because he dislikes short sellers’ methods: reports can turn assumptions or exaggerated claims into profit-seeking narratives, and he doubts a long-only market would look “terribly different.” His own 10-year stock choice is Amazon; for Dave, he predicts multiple credit products, greater primary-bank adoption, and much deeper monetization than its three-year-old checking and 10-year-old ExtraCash businesses provide today.
Verification Notes
- The host states a $4 billion IPO market cap, while Wilk later refers to a $5 billion valuation before the collapse; the digest preserves that attribution discrepancy rather than resolving it.