Upstart Holdings, Inc. (UPST): what the price requires
The current priced-in claim for Upstart Holdings, Inc. (UPST) is temporarily suppressed because the live engine record is unavailable. The dated report remains a snapshot, not a current market read.
Generated: 2026-07-14 · Exported: 2026-07-16 · Source: https://boothcheck.com/report/UPST
Headline
| Field | Value |
|---|---|
| Ticker | UPST |
| Company | Upstart Holdings, Inc. |
| Current price | $31.27/sh |
What The Price Requires (Inversion)
The assumption today's price embeds, recovered by inverting the valuation.
| Field | Value |
|---|---|
| Inversion basis | financials |
| Price-to-book | 4.08x |
The implied return on book is non-physical at this price-to-book and is suppressed as misleading. The rarity read below is the honest signal.
How unusual the bet is: extreme
| Reference | Value |
|---|---|
| vs own history | +4.96σ |
| cohort percentile (of 72 peers) | 57 |
| sustained it ~10 years at this level | 24% |
| implied end-window share | 0% |
Valuation X-Ray
Asset, earnings-power and peer-multiple models all land far below the price; ONLY the growth-DCF reaches it. The bet is durable compounding the static frames structurally cannot price (a moat/durability premium).
How the valuation models price the stock relative to the market price. Price/FV above 1.0 means the market pays more than that lens defends (expensive); at or below 1.0 the lens can defend the price.
| Family | Median price/FV | Models | Reads |
|---|---|---|---|
| Asset | 6.20x | 4 | expensive |
| Earnings | 4.71x | 2 | expensive |
| Relative | 2.16x | 4 | expensive |
| Growth | 0.69x | 1 | justifies |
Families that justify the price: Growth Families that call it expensive: Asset, Earnings, Relative
The models below discount at their own flat-beta convention rates (cost of equity 9.3%, WACC 7.6%); the inversion above states its own rate.
Per-Model Detail (n=11)
| Model | Family | FV | Price/FV | Applicable | Methodology |
|---|---|---|---|---|---|
| DCF Perpetual Growth | Growth | $0.00 | — | no | Negative/zero FCF — equity value floored at $0 |
| DCF Exit Multiple | Growth | $0.00 | — | no | Negative/zero FCF or EBITDA — equity value floored at $0 |
| Relative Valuation | Relative | $13.70 | 2.28x | yes | P/E 31x (blended: static sector reference 18x + trailing (TTM) 61x), scenarios: 24.8x / 31.0x / 37.2x (bear / base = reference held flat / bull), EV/EBITDA 26.4x |
| Simple DDM | Growth | — | — | no | — |
| Two-Stage DDM | Growth | — | — | no | — |
| Simple Excess Return | Asset | $5.51 | 5.68x | yes | BV/sh $7.57, ROE (TTM) 6.7%, ke 9.3% |
| Two-Stage Excess Return | Asset | $4.65 | 6.72x | yes | 5yr excess ROE then converge to ke=9.3% |
| Discounted Future Market Cap | Growth | $45.37 | 0.69x | yes | Rev $1.1B, growth 30% (input: historical growth; tapered), Terminal P/S: 2.1x / 2.7x / 3.2x (bear / base = today's held flat / bull, cap 12x) |
| Peter Lynch Fair Value | Relative | $4.92 | 6.36x | yes | EPS $0.41, growth 2% (input: historical EPS growth), PEG=30.67 (Overvalued) |
| Margin Trajectory | Growth | — | — | no | — |
| Earnings Power Value | Earnings | — | — | no | — |
| Residual Income | Asset | $4.53 | 6.90x | yes | BV $7.57 + 5yr PV of (ROE (TTM) 6.7% − Kₑ 9.3%) × BV; BV grows 4.4%/yr |
| Graham Number | Asset | $8.35 | 3.74x | yes | √(22.5 × EPS $0.41 × BVPS $7.57) — Graham's conservative floor |
| EV/EBITDA Relative | Relative | $0.01 | 3127.00x | yes | EBITDA $0.06B × sector EV/EBITDA 12.0x (excluded from median) |
| FCF Yield | Earnings | — | — | no | — |
| SBC-Adj FCF Yield | Earnings | — | — | no | — |
| Ben Graham Formula | Earnings | $13.23 | 2.36x | yes | EPS $0.41 × (8.5 + 2×15.0%) × (4.4 / 5.3%) |
| ROIC-Justified P/B | Asset | — | — | no | — |
| P/Sales Sector | Relative | $29.38 | 1.06x | yes | Revenue $1.14B × sector P/S 2.5x |
| PEG Fair Value | Relative | $15.37 | 2.03x | yes | EPS $0.41 × (PEG 1.5 × growth 25.0% (input: historical EPS growth)) → PE 37.5x |
| Earnings Yield | Earnings | $4.43 | 7.06x | yes | EPS $0.41 / required return 9.3% (Rf 4.3% + ERP 5.0%) |
| Funds From Operations Multiple | Relative | — | — | no | — |
| Clinical Phase NPV | Growth | — | — | no | — |
| Merton | Asset | — | — | no | — |
| V5 Mechanical | — | — | — | no | — |
Solvency
| Field | Value |
|---|---|
| Share count CAGR (dilution) | 0.4% |
Deposit/float-funded balance sheet: debt is funding, not corporate leverage, and GAAP operating cash flow follows loan flows. Net-debt, interest-coverage, and cash-burn lenses do not apply. The solvency frame for a financial is regulatory capital and payout capacity (CET1, stress buffer, dividends plus buybacks against earnings).
Bullet Takeaways
- Upstart is an AI lending marketplace, not a balance-sheet lender: it underwrites consumer loans with machine-learning models and funds them through lending partners and institutional investors, securing over $4 billion of new committed capital so far in 2026 to keep the volume flowing.
- Growth has re-accelerated hard: Q1 2026 originations of about $3.4 billion rose 61% year over year, revenue of $308 million rose 44%, and home loans grew 250%, with full-year 2026 revenue guided to roughly $1.4 billion.
- The defining risk is the cycle: the price embeds many years of uninterrupted compounding, while origination volume swings with interest rates, credit appetite, and the willingness of funding partners to buy loans, which the company names as a core dependency.
Bull Case
Capital is the raw material of a lending marketplace, and Upstart's most important recent achievement is securing more of it. A platform that does not hold most of the loans it originates lives or dies on whether funding partners will buy what it underwrites, and in 2026 Upstart locked in over $4 billion of new committed capital year to date, including a 24-month commitment and renewals from key partners, plus $1 billion of securitizations that were multiple times oversubscribed. Oversubscribed securitizations are the market voting that the loans Upstart's models produce are performing, which is the validation that matters more than any internal metric. The company also applied for a national bank charter, a move that would give it a more permanent, lower-cost funding base and reduce its dependence on the appetite of third-party buyers. How a marketplace lender sources and commits its funding is its capital allocation, and the direction here is toward more durable, more committed capital.
The operating re-acceleration is the second leg. Q1 2026 originations reached about $3.4 billion, up 61% year over year, on 425,356 loans, up 77%, with revenue of $308 million up 44% and fee revenue of $277 million up 49%. The standout was home lending, where originations grew 250% year over year, with over a quarter of home loans fully automated and an average time to close of six days. That automation is the moat in action: Upstart's model assesses applications that pass a lending partner's hard credit criteria through a pricing model, and the more of that process the AI handles end to end, the lower the cost to originate and the wider the approval funnel.
The bull thesis rests on the model improving faster than competitors can copy. Upstart reported its underwriting-model accuracy lead expanding to a wide margin above benchmark, with management framing most of the potential accuracy still unsolved, which is the company's way of saying the runway for model-driven gains is long. For a business priced on durable compounding, that is the substance of the bet: if the AI keeps approving more good borrowers at better prices than legacy credit scoring, the platform takes share across loan categories, and the home and auto expansions show it can extend beyond its personal-loan roots. The price is paying for years of that compounding, and the Q1 re-acceleration plus the committed-capital wins are the evidence the engine is running again.
Bear Case
Upstart's earnings are a leveraged bet on the credit cycle, and that is the bear's spine. Loan origination volume rises when rates are favorable, credit demand is strong, and funding partners are eager to buy; it falls hard when any of those reverse. The company is explicit that if lending partners cease or limit their participation in our marketplace or if it cannot attract new ones, its business and financial condition could be adversely affected, and that the servicing fees from loans sold to institutional investors are a material part of the model. The same automation and committed capital that power the bull case in a good cycle become the exposure in a bad one: a credit downturn that pushes loss rates above what the models predicted would chill the funding appetite exactly when volume is already softening, and a marketplace lender with thin funding is a marketplace lender that cannot originate.
The price assumes that cyclicality essentially does not happen for a very long time. At today's level the price embeds many years of uninterrupted compounding, a duration assumption that sits at the extreme end of what any business sustains, let alone a lender whose volume is rate- and credit-sensitive. The asset-based and earnings-power methods land far below the price because the company earns only a thin return on its book today, with return on equity near 6.7% and trailing earnings per share around $0.41. Only the forward-growth method reaches the price, and it does so by extrapolating the current growth rate forward. So the entire valuation rests on the durability of growth that history says is the first thing to break when the credit cycle turns.
The model risk is its own category. Upstart's edge is that its AI predicts credit outcomes better than legacy scoring, but that edge has never been tested across a full, severe downturn at the current scale and across the new categories like home and auto. If the models prove less accurate when conditions diverge sharply from the data they trained on, losses run hotter than priced, the funding partners pull back, and the flywheel reverses. The right balance-sheet frame here is funding access and capital adequacy, not corporate leverage, because the loans are funded by partners and securitizations rather than by debt UPST services. The bear is not that Upstart's technology does not work. It is that the price requires the growth to compound through cycles the model has not yet weathered, on a funding base that can tighten faster than the volume can adjust.
Valuation
The price is making a long-duration growth bet, and that is the cleanest way to describe it. Inverting today's price near $32 (as of June 27, 2026) implies the business compounds at a high rate for an extraordinarily long stretch, a duration assumption at the far extreme of what companies sustain. This is not a valuation anchored to what Upstart earns today; it is a bet on the AI-lending model taking durable share for many years, and the report isolates that bet on purpose rather than pretending trailing multiples capture it.
The methods make the structure plain. Every static family lands well below the price: the asset-based methods near $4.50 to $5.50 against a book value of $7.57 a share and a thin 6.7% return on equity, the earnings-power methods in the single digits because there is little current profit to capitalize, and the relative-multiple methods in the low-to-mid teens. Only the forward-growth method reaches the price, landing near $47 by crediting roughly 30% revenue growth tapering forward. The pattern is unambiguous: the price is a bet beyond what any backward-looking method supports, defended solely by the assumption that the growth persists. For a narrative-priced growth fintech, that is the honest read, and the trailing earnings multiple is simply the wrong anchor for a holder.
The balance-sheet question for a marketplace lender is funding access, not net debt. Upstart funds its originations through lending partners and institutional buyers rather than carrying most loans itself, so the relevant solvency frame is the committed capital behind the platform and the appetite of the securitization market. On that frame the recent news is favorable: over $4 billion of new committed capital secured year to date and oversubscribed securitizations. The peer set, other consumer-finance and fintech lenders like SoFi, Dave, UWM, and PennyMac, all live on the same funding-and-credit dynamics, so the relevant comparison is growth durability and model performance, not a single multiple. What the buyer is underwriting is many years of compounding from an AI underwriting edge, through credit cycles that edge has not yet been tested against at scale.
Catalysts
The Q1 2026 report was the catalyst that confirmed the re-acceleration. Upstart originated 425,356 loans, up 77% year over year, for about $3.4 billion in volume, up 61%, and reported total revenue of $308 million, up 44%, with fee revenue of $277 million, up 49%. Home lending was the standout, with originations up 250% year over year and more than a quarter of home loans fully automated at an average six-day close. Management guided full-year 2026 revenue to approximately $1.4 billion and fee revenue to approximately $1.3 billion.
The funding and regulatory developments are the other live catalysts. Upstart secured over $4 billion of new committed capital year to date, including a 24-month commitment and partner renewals, and completed $1 billion of securitizations that were multiple times oversubscribed, which directly addresses the platform's most important constraint. The company also submitted an application for a national bank charter, which over time could lower its funding cost and reduce dependence on third-party buyers.
The catalysts to watch are the monthly and quarterly origination volumes, which signal whether the growth is holding, the credit performance of recent vintages, since loss rates relative to the model's predictions drive funding-partner confidence, and any progress on the bank-charter application. The macro backdrop, particularly the direction of interest rates and consumer credit, is the variable with the most leverage on the volume trajectory the price depends on.
Peer Cohorts (Per Segment, With Filing Citations)
Personal Lending (sole reportable segment ~ whole company) (reported)
- SOFI (SoFi Technologies, Inc.)
- (no filing in the citation store)
- LC (LendingClub Corporation)
- (no filing in the citation store)
- ENVA (Enova International, Inc.)
- (no filing in the citation store)
- DAVE (Dave Inc./DE)
- (no filing in the citation store)
- UWMC (UWM HOLDINGS CORPORATION)
- (no filing in the citation store)
- PFSI (PennyMac Financial Services, Inc.)
- (no filing in the citation store)
- WD (Walker & Dunlop, Inc.)
- (no filing in the citation store)
Methodology Note
- Priced-in inversion: the valuation is inverted on the current price to recover the operating-income growth, duration, and steady-state margin the price embeds (ROE for financials, FFO growth for REITs).
- Valuation x-ray: the valuation models, grouped into four families (asset, earnings, relative, growth). Each model is expressed as a price/FV ratio (distance from price), not a point fair-value estimate. The spread across families is the disagreement.
- Solvency: net cash/debt, net-debt-to-NOPAT, interest coverage, and share-count CAGR from EDGAR financials (net debt / FFO and fixed-charge coverage for REITs; regulatory-capital framing for financials).
- Peer cohorts: per-segment comparables with deep-linkable SEC filing citations.
Fundamentals sourced from SEC EDGAR filings. Current price from Databento. The priced-in inversion and valuation x-ray are computed by the boothcheck engine; narrative composed by AI from the structured data.
Sources
Upstart Q1 2026 earnings release, May 2026 · Upstart FY2024 10-K