Thesis
SoFi Technologies is a legitimate developing compounder at an early but identifiable inflection point. The company crossed the $1 billion quarterly revenue threshold in Q4 2025, achieved its ninth consecutive quarter of GAAP profitability, and is executing a three-segment platform strategy — Lending, Financial Services, and Technology Platform — that creates genuine cross-subsidization economics.
With 13.7 million members growing at 35% YoY, a deposit base of $37.5B (97% FDIC-insured primary accounts), and fee-based revenue now annualizing at $1.8B, the business is structurally differentiated from pure-play neobanks.
The central thesis: SoFi’s current 18% operating margin grossly understates its 5-year potential. The Financial Services segment inflected to 51% contribution margins in Q4 2025 — from near-zero in 2023 — on an asset-light model. The Technology Platform (Galileo + Technisys) is an overlooked embedded-finance infrastructure asset with $1B+ revenue potential. Management guides for 30%+ CAGR in adjusted net revenue and 38–42% EPS CAGR through 2028.
5-year base case target: $60–75B market cap vs. $24.1B today. Implied return of 150–200% over 5 years — roughly 20–24% annualized — if the thesis executes.
Three-Segment Platform: The Financial Flywheel
Segment 1: Lending (~51% of FY2025 Revenue, ~$1.84B)
Products: Personal loans, student loan refinancing, home loans. The original SoFi franchise and still the dominant revenue generator. The key structural shift: SoFi is migrating from an on-balance-sheet model to a capital-light ‘Loan Platform Business’ (LPB), where loans are originated for fee income on behalf of third-party institutions. The LPB originated $11B in personal loans for third parties in FY2025, generating $576M in fees — a shift from NII-dependent to fee-based, rate-agnostic revenue.
Average FICO of personal loan borrowers: 746. Average income: $158,000. This is demonstrably a prime borrower base.
Segment 2: Financial Services (~35% of FY2025 Revenue, ~$1.27B)
Products: SoFi Money (checking/savings), SoFi Relay (financial aggregator), SoFi Invest (brokerage), credit cards, insurance. The fastest-growing segment — Q4 2025 revenue grew 78% YoY to $457M. Revenue per Financial Services product reached $104 annualized in Q4 2025, up 29% YoY. Contribution margins reached 51% in Q4 2025, up from near-breakeven in early 2024.
Total deposits: $37.5B, with 97% from primary (direct deposit) accounts. This is a critical structural advantage — primary accounts represent sticky, low-cost funding. The SoFi Plus subscription ($10/month launching March 2026) adds a recurring SaaS-like revenue layer.
Segment 3: Technology Platform (~14% of FY2025 Revenue, ~$450M)
Galileo (payments/API infrastructure) and Technisys (cloud core banking). The ‘AWS of fintech.’ FY2025 record revenue of $450M. The segment lost a major client in 2025, resulting in a 23% account decline — a headwind. But the replacement pipeline (Treasury Direct Express, Southwest Airlines, Wyndham, two additional large consumer brands) suggests 2026 recovery.
Galileo targets $1B in revenue. At $450M today, the implied trajectory requires acceleration; management guidance of 20%+ normalized growth implies $540–570M for 2026.
The Three Moat Sources
Banking Charter + Deposit Franchise (STRENGTHENING)
SoFi’s national bank charter (January 2022) is a structural moat most fintech competitors lack. The charter allows FDIC-insured deposits, eliminating reliance on high-cost wholesale funding. The average rate paid on deposits in Q4 2025 was 181bps lower than warehouse facilities — translating to ~$680M in annualized interest expense savings. This is permanent. Competitors like Robinhood and Chime (no charter) must either pay wholesale rates or partner with banks — neither matches SoFi’s economics.
Galileo Technology Infrastructure (MODERATE, EARLY)
Financial institutions that build card programs on Galileo face $2–10M+ integration costs and 12–24 months of re-implementation to switch providers. The Treasury’s Direct Express partnership (3.4M beneficiaries), Southwest Airlines, and Wyndham Hotels all represent multi-year contracts with high operational switching costs. Technisys adds core banking conversion capability — migrating a bank’s core is a decade-long project. This moat is currently undervalued because the segment is sub-scale, but it represents the highest-quality recurring revenue in the portfolio.
Financial Services Productivity Loop (EMERGING FLYWHEEL)
Acquire members via high-margin lending → convert to deposit accounts → cross-sell investing, credit, and insurance. Each product added increases data richness, reduces default risk, and increases lifetime value. Cross-buy rate reached 40% in Q4 2025 — 40% of new products opened by existing members, compared to ~25% for traditional banks.
NuBank: The Most Instructive Comparable
NuBank went from -8% operating margins to ~30% in 3 years at comparable revenue scale. The key driver was deposit cost advantage flowing through to the bottom line as the loan book grew on cheap funding. SoFi is executing the same playbook at 5x the scale.
LendingClub is the charter-conversion analog: bank charter obtained 2021, margin trajectory from deeply negative to ~25% operating margin in 2 years by replacing warehouse funding with deposits. SoFi is doing this at 5x the scale.
Where the Market May Be Wrong
Error 1: Linear Margin Modeling. The Financial Services segment’s operating leverage is nonlinear — contribution margins exploded from ~10% in Q1 2024 to 51% in Q4 2025 in 6 quarters. Sell-side models interpolating linearly from 18% margin to 30% over 5 years are likely 3–4 years late on the margin inflection timeline. The correct model: blended EBITDA margin goes from 31% (Q4 2025) to 38–42% by FY2027.
Error 2: Technology Platform Undervalued. No sell-side model values the Technology Platform separately from the consumer bank. At $450M in revenue growing 20%+ and targeting $1B, on a BaaS multiple of 6–8x revenue, the Tech Platform alone is worth $2.7–4.0B — 11–17% of SoFi’s current market cap. Investors who recognize this get a ‘double unlock’: earnings growth and multiple re-rating as the segment is recognized as infrastructure.
Error 3: Credit Risk Overestimated. Consensus applies subprime fintech comps (Upstart, Avant) to a borrower base that earns $158K, has FICO 746, and uses SoFi as a primary banking relationship. The appropriate comp is Marcus/Goldman or Ally Financial, where prime borrower charge-offs in recession topped out at 3.5–4.5% — materially below 6–8% scenarios embedded in some bear models.
Key Risks
Credit cycle deterioration (35–40% probability): At 5% charge-off rate (vs. 2.8% baseline), incremental $320–400M annual provision expense impairs net income 50–70%.
Credit card interest rate cap (20–25% probability): The January 2026 presidential proposal to cap credit card interest at 10% would impair ~$100–150M of FS segment revenue. Partial mitigant: a cap would drive credit card debt holders toward SoFi personal loans — a potential net positive for lending originations.
Technology Platform client concentration (25–30% probability): The segment already demonstrated vulnerability by losing 39M accounts (23% decline) from a single client exit.
Dilution from future capital raises (25–30% probability): SoFi raised $3.2B in equity in 2025 (two offerings). Current Tier 1 ratio of 22.9% (vs. 10.5% minimum) provides 3–4 years of buffer, but management has demonstrated willingness to use equity as currency.
Valuation
| Scenario | Probability | Revenue (Yr5) | Op Margin | Exit P/E | Price |
|---|---|---|---|---|---|
| Bull | 25% | $9B | 37% | 25x | ~$32 |
| Base | 40% | $7.5B | 30% | 22x | ~$24 |
| Bear | 35% | $5.5B | 20% | 15x | ~$8.50 |
Current price of $18.90 reflects approximately 35% bear probability. If our assessment of 20% bear probability is correct, the stock should trade closer to $22–25 — 16–32% above current levels on a probability-weighted basis.
The thesis is not a 10-bagger story. It is an attractive risk-adjusted setup for a business crossing from growth-stage into compounding, with a banking charter that permanently re-prices its liability structure.
Prepared by Consti Ertel | March 8, 2026 | For internal research use only. Not investment advice.