Venture debt is a senior loan, usually paired with warrants, that venture-backed startups use to extend runway between equity rounds. US venture-backed companies borrowed $81.7 billion through September 2026, already above any full year in PitchBook's data going back to 2016. It works best right after a priced round. It hurts when covenants trip before the next raise lands.
That record total makes borrowing look like the default move this year, and for a few hundred well-funded companies it is. For most founders the question is narrower: will a loan buy enough time to reach a better valuation, and can the business survive the terms if the plan slips? The answer starts with how lenders think about repayment.
Venture debt is a loan underwritten on your next round
Venture debt lenders are betting on your investors more than on your cash flow. In their study of the market, finance researchers Jesse Davis, Adair Morse and Xinxin Wang describe typical loans of $1 million to $10 million, interest rates of 10 to 15 percent, terms of 24 to 36 months, and warrants worth roughly 5 to 15 percent of the loan size (Davis, Morse and Wang). They note that the source of repayment is usually the next equity round, which is why lenders track who sits on your cap table and how much they have left to deploy.
The same NBER working paper found that early-stage firms use venture debt in one-third of financing rounds (NBER). The loan sits senior to every share class, so it gets repaid first in a sale or wind-down. That seniority separates it from the instruments founders use before a priced round, which convert into equity. Our guide to SAFEs and convertible notes covers those.
The 2026 venture debt market is big and top-heavy
The headline total overstates what a typical startup can borrow. The Q3 2026 PitchBook-NVCA Venture Monitor, released October 7, reports that VC-backed companies took on $81.7 billion in debt through September 30, across 499 deals (PitchBook-NVCA Venture Monitor). Much of it went to a handful of giants. By midyear the count stood at just 280 loans, and a single $20 billion SpaceX refinancing did much of the work (SiliconANGLE).
The median deal tells the more useful story. A review produced by lender Runway Growth Capital with PitchBook data found that 2025 deal volume held steady at roughly 1,000 transactions, the median loan rose to $5.5 million, and the 75th percentile reached $27.7 million (Runway Growth Capital and PitchBook). Equity looks just as concentrated. We covered how AI megarounds skew the funding data, and the same pattern explains why a founder's term sheet rarely resembles the market average. Our read of the Q3 2026 venture funding numbers shows the equity side of that split.
Venture debt vs equity comes down to dilution math
Borrowing makes sense when the dilution it saves is worth more than the interest and warrants it costs. Take a hypothetical company valued at $40 million post-money that needs $5 million. Raising it as equity sells 12.5 percent of the company. Borrowing it at 12 percent costs about $600,000 a year in interest while the full balance is outstanding, plus warrants. At 10 percent coverage, those warrants are worth $500,000, or roughly 1.25 percent of the company at the same price.
The research backs the logic. Davis, Morse and Wang found that dilution predicts venture debt issuance, and that venture debt lowered the closure hazard by 1.6 to 4.4 percent and raised successful exits by 4.3 to 5.3 percent (NBER). The catch is that interest is paid in cash every month, from the runway the loan was meant to extend. Model both paths over 24 months before you sign, and stress-test the version where the next round slips by two quarters. Our funding and incubation team can pressure-test that model with you.
Venture debt terms decide the real cost of the loan
The interest rate is the least negotiable line on a venture debt term sheet, and rarely the most important. Public lender filings show the structure. Hercules Capital, among the largest venture lenders, reported that 97.8 percent of its debt portfolio carried floating rates with floors and 86.8 percent was first lien senior secured, with a GAAP effective yield of 13.4 percent in Q2 2026 (Hercules Capital SEC filing). Floors protect the lender when rates fall. First lien means the lender holds a claim on your assets.
The quieter clauses matter more. Hercules says part of its unfunded commitments depend on borrowers hitting milestones, and that its credit agreements let it decline to fund previously committed tranches after materially adverse events (Hercules Capital). Read that as a warning: an undrawn tranche is an option the lender holds. Negotiate the material adverse change definition, minimum cash covenants, prepayment fees and the warrant strike price. Lenders increasingly underwrite revenue quality, so a strong net revenue retention figure is real leverage at the table.
Venture debt lenders after SVB need the same diligence as investors
Silicon Valley Bank's failure showed founders that the lender itself is a risk. PitchBook reported that SVB's venture debt portfolio made up 60 to 70 percent of the entire market when it collapsed on March 10, 2023, and that the bank held accounts for almost 50 percent of VC-backed tech and life science companies in 2022 (PitchBook). Companies that kept their deposits and their loan at one bank discovered how tightly the two were tied.
Lenders now fall into two broad camps. Banks tend to charge less and impose more financial covenants, while debt funds charge higher rates, use fewer covenants and ask for more warrants, according to the Davis, Morse and Wang study (Davis, Morse and Wang). Ask any lender how it behaved with borrowers who missed plan during 2023, and call two founders from its portfolio. Hercules, for one, says it lowers the credit grade of borrowers as they approach their next equity need within 9 to 14 months (Hercules Capital). Your lender is watching your runway as closely as your board is.
Two practical rules follow. First, read the deposit clause. If a lender asks you to keep most of your cash with it, that recreates the single point of failure that hurt SVB borrowers, so negotiate a carve-out for a second operating account. Second, ask what triggers a lender's right to call the loan, and get the answer in writing. A lender that has worked through downturns with its borrowers will have a clear, specific reply.
Venture debt works best for startups with a clear next milestone
The best time to raise venture debt is right after an equity round closes, when cash is high and the lender can see your investors' commitment. It fits a company with at least 12 months of runway without the loan, a milestone that should lift the next valuation, and revenue predictable enough to carry interest payments. It fits poorly when the loan is the only thing standing between the company and a shutdown. In that case a lender's covenants become the board.
Lender-sponsored data argues the tool is going mainstream. The Runway and PitchBook review found that venture debt-backed companies accounted for 37 percent of 2025 exit value and 18 percent of exit count (Runway Growth Capital and PitchBook). Treat that as a signal of adoption, since the report comes from a lender. Pair the loan with a fundraising plan you would run anyway, like the one in our startup fundraising guide. If you want a second set of eyes on a term sheet or a runway model, talk to our team.
