Pattern-Recognition Startup Underwriting
Assess young companies through investors, markets, teams, and recurring patterns.
- Difficulty
- Advanced
- Time to result
- ~months to results
- Steps
- 5
- Confidence
- 94%
Traditional underwriting performs poorly when a startup has little revenue history, limited collateral, and uncertain cash flow. This method substitutes a structured collection of ecosystem and company signals: the investors supporting the business, the market being pursued, the management team, the financing stage, and patterns observed across comparable companies. Those signals are then combined with the lender’s contractual position and expected recovery if the company fails. Repetition is essential because the method becomes safer as the institution observes more companies, financing cycles, and failure modes. The output is not certainty about startup success; it is a risk-adjusted lending decision that recognizes both the company’s potential and the senior lender’s downside protection.
Origin
Greg Becker described how Silicon Valley Bank adapted its startup-lending capability over decades of financing technology and life-science companies. Extracted from Masters in Business.
Core principles
- 01Young companies require different evidence than established borrowers.
- 02Repeated exposure turns qualitative signals into recognizable patterns.
- 03Investor quality, market potential, and management capability must be assessed together.
- 04Debt risk differs from equity risk because senior lenders have downside protection.
How to run it
- 1
Classify the opportunity
Establish the company’s sector, financing stage, intended market, and current sources of capital. Determine whether it fits an area where the lender has meaningful comparative experience.
Pro tip Use narrowly relevant comparisons rather than treating all startups as one asset class.
Watch out A familiar startup label does not make an unfamiliar market understandable.
- 2
Assess the backers
Examine who the investors are, their relevant track record, and their capacity to continue supporting the company. Treat investor quality as one signal rather than a guarantee.
Pro tip Look for evidence of follow-on capacity and sector expertise.
Watch out Prestigious investors can still back companies that fail.
- 3
Evaluate market and management
Estimate the opportunity if the company succeeds, then assess whether its management team has the capability and drive to pursue it. Balance identifiable obstacles with the scale of the possible outcome.
Pro tip Ask both why the plan might fail and what becomes possible if it works.
Watch out Optimism without operating evidence can disguise weak execution.
- 4
Match recurring patterns
Compare the company with prior borrowers and relevant public or private peers. Identify patterns associated with successful financing, distress, and recoveries.
Pro tip Record outcomes so institutional pattern recognition improves over time.
Watch out Do not force a novel business into an irrelevant historical template.
- 5
Underwrite the downside
Model repayment, seniority, collateral, runway, and likely recovery during a wind-down. Price and structure the loan for lending returns rather than venture-equity returns.
Pro tip Stress-test the case with no warrant gains or additional financing.
Watch out Strong markets can make warrant gains and low losses look permanently easy.
In the wild
A specialist lender reviews a young cybersecurity company with limited revenue. It examines the venture sponsors, compares growth and retention with relevant software companies, assesses the founders, and models recovery if the next funding round fails. The lender offers a smaller facility tied to milestones rather than rejecting the company solely for lacking a long financial record.
→ The company receives appropriately structured capital while the lender limits downside exposure.
Becker describes evaluating the size of the opportunity, the investors, and the management team rather than stopping at a list of reasons the unconventional model might fail. SVB participated as a lender rather than the sole equity investor, preserving a different risk-and-return profile.
→ The lender supported a high-potential company without assuming the same downside as its equity investors.
Common mistakes
Using only traditional financial history
A new company may have too little operating history for conventional ratios to carry the decision. Rejecting it automatically ignores other relevant evidence.
Confusing debt with equity
Lenders have capped upside, contractual seniority, and possible recovery during a wind-down. Their underwriting should not depend on finding a few enormous equity winners.
Trusting patterns without adapting
Patterns are useful only when updated for new markets, financing sources, and business models.
Is it for you?
Best for
It is best for specialist lenders and investors repeatedly evaluating venture-backed technology or life-science companies.
Not ideal for
It is not ideal for occasional generalist lenders without enough sector exposure to develop reliable pattern recognition.
From the transcript
“it's really about pattern recognition. It's about understanding who the investors are, it's understanding what market they're in.”
“understanding the management team and there and their dry for success. All those things are ingredients that really gives you that view of the potential…”
“we as a senior lender in these loans, we get repaid first when you when a company goes out of business, we get repaid as…”
From the episode
Greg Becker on the Innovation Business (Podcast)
Greg Becker