Key Takeaways

  • Max Levchin incubated Affirm inside HVF a decade after leaving PayPal, targeting consumer credit after personal frustration with traditional credit bureaus.
  • Affirm reached over $50 billion in gross merchandise volume by eliminating late fees, deferred interest, and compounding balances in favor of transaction-level underwriting.
  • Building honest financial products required recruiting mathematical researchers who would otherwise design high-frequency trading models for quantitative hedge funds.
  • Incumbent banks avoid fixed-duration, zero-late-fee consumer loans because their profit models depend directly on customer penalties and revolving debt.
  • Levchin evaluates every new venture through Max Levchin's HVF Startup Evaluation Framework.

The Max Levchin HVF Startup Evaluation Framework

Fraser Christie notes that “Max talks about this framework he has in his head of starting businesses of the problem to be solved has to be hard, valuable and fun.” As Patrick O'Shaughnessy points out, “HVF was the name of his incubation which is how a firm got started too.” Levchin uses three criteria to filter potential ventures:

  • Component 1: Hard Problems (Hard): The problem to be solved must be technically and operationally difficult. Leaning into hard problems, such as underwriting unsecured consumer loans over long durations without late fees or compounding interest, creates defensible, structural competitive moats.
  • Component 2: Valuable Economic Impact (Valuable): Solving the problem must generate massive economic value for both consumers and merchants by fixing deeply flawed, extractive legacy models.
  • Component 3: Fun Technical Complexity (Fun): The mission and complexity must be intellectually engaging enough to attract world-class engineering, cryptography, and mathematical talent.

When This Works (and When It Doesn't)

Levchin used this filter to evaluate incubation concepts at HVF, screening out superficial software features in favor of high-friction opportunities that banks deliberately avoid. “A decade after he left PayPal, Max started a firm with a mission that still stands today, which is to deliver honest financial products that improve lives,” Christie explains. Eliminating late fees meant Affirm had to calculate credit risk accurately at the point of sale. That operational barrier protected the company from copycats. Christie notes that “Max wanted to set out securing the type of mathematical talent that otherwise ends up in scientific research or hedge funds building models to make money in split-second decisions.”

The model fails when early teams lack the technical credibility or balance sheet access required to survive prolonged friction. Tackling a structurally hard problem in regulated finance without deep domain expertise or capital markets support creates operational paralysis. If you cannot recruit top mathematical talent or fund loans through capital cycles, high friction will break your company long before you build a defensible moat.

What to Do With This

Run your current startup backlog through Levchin's three filters this week. Sit down with your co-founder tomorrow morning and score your core product thesis against these tests:

First, identify the hardest technical constraint you currently avoid. If your product relies on simple wrappers around existing APIs, you have built a feature that an incumbent can copy in a quarter. Find the high-friction operational bottleneck in your industry, such as custom transaction-level risk pricing, that competitors run away from because it hurts short-term margin.

Second, audit your revenue model for extractive penalties. If your unit economics depend on customer mistakes, missed deadlines, or confusing pricing tiers, rewrite the contract. Align merchant conversion directly with transparent customer outcomes.

Third, test whether your core problem attracts elite technical specialists. “Max's personal DNA of leaning into hard problems you see throughout the whole organization,” Christie says. If your engineering pitch sounds like routine CRUD development, increase the mathematical or architectural ambition until top researchers leave research labs to solve it.