Final investment rating: 6.5/10 - selective accumulation, with valuation discipline.
I like SoFi’s business. It has developed a broad financial platform with growing members, improving cross-selling and real earnings. However, its lending business carries credit and capital risks, and its valuation remains sensitive to future growth assumptions.
My simplified base DCF estimates US$12.91 per share. A 50/50 blend of that model and Alpha Spread’s DCF valuation produces a reference value of approximately US$15.83 per share, using unrounded calculations.
Compared with the reference price of US$15.80 during the US session on 6 October 2026, SoFi appears approximately fairly valued under the blended estimate. That leaves little margin of safety. The blend is a secondary reference, rather than a replacement for examining each model’s assumptions.
| Assessment | My view |
|---|---|
| Business quality | 8/10 |
| Investment attractiveness at the reference price | 6.5/10 |
| My simplified base DCF | US$12.91 |
| Blended fair-value estimate | US$15.83 |
| Price 20% below the blended estimate | US$12.66 |
| Approach | Accumulate selectively, considering valuation, portfolio allocation and business performance |
What does SoFi actually do?
SoFi Technologies trades on NASDAQ under ticker SOFI. It operates a digital financial platform that combines consumer banking and lending with financial technology sold to other businesses. Company overview.
| Business | What customers use it for | Main economic engine |
|---|---|---|
| Lending | Personal loans, student loans and home financing | Interest income, origination and servicing economics |
| Financial services | Banking, cards, investing and other financial products | Deposit funding, interchange, fees and deeper customer relationships |
| Technology platform | Payment processing and banking infrastructure, including Galileo and Technisys | Enterprise contracts, platform and transaction fees |
The investment thesis is that one customer relationship can support several products. A member might start with a loan or savings account and later use investing, cards and other services. If that happens efficiently, customer acquisition spending produces more revenue over time.
This gives SoFi more routes to growth than a lender dependent on a single loan product. It does not eliminate the risks of lending.
What the latest results show
| Q2 2026 metric | Reported result |
|---|---|
| Adjusted net revenue | US$1.206B; +40% YoY |
| Adjusted EBITDA | US$357.8M; +44% YoY |
| GAAP net income | US$156.6M |
| Diluted EPS | US$0.12 |
| Members | 15.8M; +35% YoY |
| Products per member | 1.54 |
SoFi also guided to US$4.75–4.85 billion of full-year adjusted revenue and approximately US$825 million of adjusted net income. Adjusted results and guidance are not the same as GAAP earnings or distributable cash.
The important question is whether earnings can keep growing after credit costs, marketing, technology investment and dilution. A good revenue quarter does not by itself establish an attractive purchase price.
Why I like the business
If you would like to see SOFI financials: Q2 earnings release
1. Cross-selling can improve customer economics
In Q2, 51% of new products were opened by existing members. Products per member reached 1.54.
I would say this is a great start, providing room for deeper relationships, and the cross-buy figure gives evidence that existing users are adopting more products.
2. Deposit funding supports the lending business
SoFi reported US$45.5 billion of deposits at quarter-end.
Deposits can provide a more dependable funding base than relying entirely on wholesale facilities. The benefit depends on what SoFi pays depositors, what its assets earn, and how well it manages liquidity. Deposits belong to customers; they are not cash shareholders can freely withdraw.
3. More fee-based activity can reduce dependence on held loans
Originating or servicing loans for third parties and selling financial infrastructure can grow revenue without requiring SoFi to retain every loan. My long-term preference is for growth that produces good returns with manageable capital requirements.
I would watch the balance between loans held, loans sold and loans facilitated for partners. Moving risk off the balance sheet can help capital efficiency, but partner demand and loan-sale pricing still matter.
4. Profitability is now a meaningful part of the thesis
The historical run from Q4 2023 through Q2 2026 covers eleven consecutive GAAP-profitable quarters. That is useful evidence of progress, but future profitability must be earned again as credit conditions and investment spending change.
Credit quality: the most important risk to monitor
Management reported 62% principal paydown and 6.8% net cumulative losses for the Q1 2020–Q1 2026 personal-loan cohort, against its 7–8% lifetime loss tolerance. Its estimated annualised charge-off rate excluding late-stage delinquent loan sales was approximately 3.7%; 90-day delinquency rates were consistent with the prior year. Q2 credit discussion.
These measures describe different things. Cumulative cohort losses, annualised charge-offs and delinquency rates should not be compared as if they use the same denominator or time period. Management’s tolerance is an underwriting objective, not a guaranteed ceiling.
The current credit evidence is encouraging, but rapid loan growth can make percentage measures look better before new loans have fully seasoned. I would track older vintages, recoveries and delinquent-loan sales alongside the headline rates.
Personal loans are particularly exposed to unemployment and borrower cash-flow stress. In a recession, losses can rise while demand, investor appetite and valuation multiples weaken together.
Interest rates: a mixed effect
The Federal Reserve raised its target range to 3.75–4.00% on 16 September 2026. Federal Reserve statement.
I would model several rate paths instead of assuming a fixed number of additional hikes. Higher rates can pressure refinancing demand, repayment affordability and equity valuations. They can also increase asset yields. The net effect depends on funding costs, loan repricing, customer behaviour and credit performance.
Lower rates can support refinancing and reduce funding pressure, but they can also lower yields and reserve income. The relevant question is how the whole balance sheet and fee business respond.
SoFiUSD and Mastercard, explained simply
SoFiUSD is a dollar-backed digital token intended to maintain a value of one US dollar. Its role in Mastercard settlement concerns the movement of money between financial institutions after card transactions, rather than requiring a customer to use cryptocurrency at the checkout.
SoFi announced live stablecoin settlement across its Mastercard card programme in September. Its full programme was expected to represent more than US$25 billion of annualised volume. This is payment volume, not revenue or stablecoin supply. Settlement announcement.
For example, a customer pays by card, Mastercard processes the transaction, and the institutions can settle their obligations using SoFiUSD. The same token can circulate repeatedly, so substantial settlement volume does not require an equal amount of outstanding tokens.
How could this help SoFi?
| Potential benefit | Why it might matter | Evidence I would want |
|---|---|---|
| Faster settlement | Better timing and liquidity management | Measurable operating or treasury savings |
| Reserve economics | Income on eligible reserve assets | Net yield after reserve, compliance and distribution costs |
| Enterprise adoption | More infrastructure customers | Third-party usage, contracts and associated revenue |
| Broader payments services | New cross-border and business-payment use cases | Commercial adoption beyond the initial card programme |
Galileo distribution makes the enterprise opportunity interesting, because external financial institutions could become users. That is a potential growth route rather than demonstrated large-scale profit today. SoFi–Mastercard partnership.
An illustration of reserve economics is US$300M × 4% = US$12M of gross annual income. This assumes an interest-earning reserve balance and is not a claim about current supply, actual reserve composition or net profit. At US$5 billion, the same assumed yield would generate US$200 million gross, but reaching that scale is speculative.
I treat SoFiUSD as potential upside within the broader business. I assign no separate premium in the DCF. Stablecoin reserves are backing obligations, and adding their balances to shareholder value would be inappropriate.
Risks that keep my rating below a strong buy
| Risk | Why it matters |
|---|---|
| Personal-loan credit cycle | Defaults can rise after the portfolio expands |
| Capital requirements | Lending growth can consume profits rather than release cash |
| Deposit competition | Higher funding costs can reduce spreads |
| Technology execution | Enterprise growth must support the diversified-platform thesis |
| Dilution and stock compensation | Company profit can grow faster than each share’s economic value |
| Loan valuation and sales | Fair-value assumptions and sale prices affect reported results |
| Stablecoin execution and regulation | Adoption may be slower or less profitable than expected |
| Valuation | Good results can still disappoint if the price anticipates more |
Technology Platform revenue fell 23% YoY in Q2. Q2 earnings release. I would want enterprise performance to recover before assuming every part of the business is growing strongly.
A flat share price does not mean a rally is overdue. Likewise, a P/E premium to mature lenders only makes sense if future growth and returns justify it. Different lenders and technology businesses are imperfect comparisons.
My simplified DCF
For a bank-led business, I discount estimated cash available to equity holders after retaining capital for growth:
Revenue × normalised net profit margin = estimated net income
Net income × (1 − retention rate) = estimated shareholder cash
Shareholder cash ÷ projected diluted shares = cash per shareThis is a top-down equity DCF, not a complete regulatory-capital or accounting cash-flow forecast. Retention is a modelling proxy, not SoFi’s dividend policy. Higher actual capital needs would reduce value.
Base assumptions
| Input | Assumption |
|---|---|
| 2026 revenue starting point | US$4.80B, midpoint of adjusted guidance |
| Forecast | 2027–2036, followed by continuing value |
| Annual revenue growth | 25%, 23%, 20%, 18%, 15%, 12%, 10%, 8%, 6%, 4% |
| Annual net profit margins | 18%, 19%, 20%, 21%, then 22% |
| Retained earnings | 50% for five years; 30% for five years |
| Initial diluted shares | 1.38B, assumed modelling base |
| Annual net dilution | 1% |
| Required equity return | 11.5% |
| Terminal aggregate earnings growth | 3% |
| Terminal ROE | 18% |
Stock compensation remains an expense in the assumed margins. Terminal retention is 3% ÷ 18% = 16.67%, allowing for the capital needed to sustain growth. Continuing dilution reduces terminal per-share growth to approximately 1.98%.
What the base case requires
| Year | Revenue, US$B | Net profit margin | Net profit, US$B | Cash after retention, US$B |
|---|---|---|---|---|
| 2027 | 6.00 | 18% | 1.08 | 0.54 |
| 2031 | 12.02 | 22% | 2.64 | 1.32 |
| 2036 | 17.63 | 22% | 3.88 | 2.71 |
The change in retention after 2031 increases cash availability. It is a model transition, not a forecast of a particular corporate announcement.
| Present-value component | Per share |
|---|---|
| Forecast shareholder cash | US$5.46 |
| Continuing value | US$7.44 |
| Base estimate, using unrounded calculations | US$12.91 |
Approximately 58% of the value comes from the continuing period. The model uses year-end cash flows with a 0.25-year initial stub, and omits the remaining 2026 cash flow. No separate book-value, deposit-cash or stablecoin premium is added.
Bear, base and bull outcomes
| Scenario | Estimated value per share | Interpretation |
|---|---|---|
| Bear | US$4.36 | Slower growth, weaker cash generation and higher required return |
| Base | US$12.91 | Solid growth, improving profitability and ongoing investment |
| Bull | US$26.83 | Strong execution, better economics and lower perceived risk |
The bear case ends at 18% net margins and uses a 13.5% required return; the bull case ends at 25% margins and uses 10.5%. They also vary growth, retention and dilution. These are scenarios, not probability bands.
Required-return sensitivity
| Required return | Base value per share |
|---|---|
| 10% | US$16.10 |
| 11.5% | US$12.91 |
| 13% | US$10.63 |
Other base assumptions stay unchanged. The choice of required return is a major reason DCF values differ.
What other websites estimate
| Provider | Published estimate | Method |
|---|---|---|
| Alpha Spread | About US$18.76 | Bank-template DCF with a 3× revenue exit multiple |
| Simply Wall St | US$12.21 | Intrinsic valuation; its SoFi analysis uses excess returns |
| ValueInvesting.io | US$12.33 | Peter Lynch fair value; not DCF |
These are retrieved snapshots from the October 2026 research, not harmonised models. Alpha Spread’s retrieved current model used an 8.8% discount rate, below my 11.5%. Its terminal method also differs. Method details.
I would compare assumptions rather than average these values. None is authoritative, including mine.
What would change my rating?
| Raise my rating | Lower my rating |
|---|---|
| Strong growth converts into higher earnings per share | Credit deterioration outpaces revenue growth |
| Cross-selling improves active customer economics | Products increase without profitable engagement |
| Technology performance recovers | Enterprise weakness persists |
| Lending growth becomes more capital efficient | Capital needs force substantial new equity |
| Stablecoin adoption produces visible economics | Partnerships remain financially immaterial |
| Price offers a larger discount | Valuation rises faster than earning power |
My final rating remains 6.5/10 at approximately US$15.80. SoFi is a business I want to keep researching, with a quality assessment of 8/10. The price-sensitive investment rating is lower because its growth opportunity must be balanced against lending risk, capital retention and valuation.
Research date: 7 October 2026. Reference price: an intraday observation on 6 October, not a closing price. Forecasts and ratings are my analytical assumptions; they are not promises of returns. Sources are linked beside the relevant claims.