SOFI (Trans): Raises FY guide; 2027 rev. seen at $6.4bn

Compiled by Dolphin Research: $SoFi Tech(SOFI.US) Q3 2026 earnings call Trans

I. Key earnings takeaways

1. Raised 2026 full-year guide (originally assumed two cuts; the Fed has hiked once and is expected to hike two more times this year)

a. Adj. net revenue of $5.04bn, +40% YoY, above the prior $4.75bn–$4.85bn range. It is $385mn higher than the initial guide at the start of the year.

b. Adj. EBITDA of $1.66bn (margin 33%, prior $1.60bn). Adj. net income of $840mn (margin 17%, prior $825mn), and EPS of $0.61 (prior $0.60).

c. Implied Q4: revenue of about $1.375bn (+36% YoY). EBITDA of $492mn, net income of $276mn (net margin 20%), and EPS of $0.19.

d. Q4 by segment: Lending $785mn, Financial Services $550mn, and Technology Platform $90mn. Corporate line roughly flat vs. Q3.

2. Initial 2027 guide (provided one quarter earlier than usual)

a. Adj. net revenue of $6.4bn, +27% YoY. EBITDA of $2.24bn with ~35% margin. Net income of $1.28bn with ~20% margin, and EPS of $0.90.

b. Assumptions: two rate hikes during the year and an effective tax rate of 24%.

c. Plan to fund growth entirely with existing excess capital and internally generated capital. No need for, and none assumed, external financing.

3. Rate shocks and hedging

a. End-of-quarter fair values marked at 103.3% for personal loans (QoQ -140bps), 101.9% for student loans (-240bps), and 101.5% for home loans (-330bps. The primary driver was higher discount rates.

b. Aggregate fair value losses exceeded $800mn, while hedging gains were about $734mn. Hedges covered base-rate moves, with residual impact from non-hedged input factors.

4. Capital, returns, and balance sheet

a. Total capital ratio at 17% (QoQ -180bps), well above the 10.5% regulatory floor and internal stress buffers. The target is to gradually normalize to the mid-to-low teens.

b. Cash revenue was $1.4bn, including net interest income of $927mn, marking the fourth straight quarter above $1bn. ROTCE was about 10%, with a long-term goal of 20%–30%.

c. Deposits rose $5.8bn QoQ to $51.3bn. NIM was 6.14% (QoQ +16bps) and is expected to remain above 5% for the foreseeable future, and tangible book value per share was $7.48 (+25% YoY).

II. Earnings call details

2.1 Key management commentary

1. Operating framework shift

a. Starting this quarter, management frames the biz. around three drivers: member count, products per member (i.e., total products), and revenue per product. Resource allocation follows the same three pillars.

b. Rule of 40 score was 79% (revenue growth 44% plus EBITDA margin 34%), meeting the target for 20 straight quarters.

2. Members and products

a. Members rose 35% YoY to 17.1mn, with a record 1.2mn net adds in the quarter. Total products increased 44% YoY to 26.7mn, with a record 2.2mn net adds.

b. Products per member reached 1.56 (vs. 1.54 last quarter and 1.47 a year ago). Fifty-one percent of new products came from existing members, and 37% of members use two or more products (vs. 33% last year).

c. Eighty-seven percent of products were Financial Services rather than Lending. Fee-based revenue was $571mn, accounting for 42% (39% in Q2), with a long-term goal of 50%.

3. Lending

a. Net interest income was $670mn (+57% YoY), and loan origination fees were $271mn (2.6x YoY). Total loan originations hit a record $16.9bn.

b. Personal loan originations were $12.6bn, a record, with $9.0bn retained on balance sheet. Risk-adjusted ROA was about 7.6% = weighted coupon 12.8% + fee yield 1.4% − funding cost 3.1% − annualized losses 3.4%.

c. Student loan originations were $2.9bn (+96% YoY), and home loans were $1.4bn (+44% YoY). Lending products reached 3.4mn (+39% YoY), with annualized revenue per product at $942 (vs. $920 last quarter and $782 a year ago).

d. Credit: total personal loan annualized net charge-off rate was 3.4% (YoY -80bps, QoQ -30bps), with 90+ DPD at 42bps. Student loan net charge-offs were 63bps, with 90+ DPD at 13bps.

4. LPB (Loan Platform Biz.)

a. Approximately $3.8bn was originated for partners (vs. $3.1bn in Q2), generating $153mn revenue (vs. $143mn in Q2). The platform transferred out its first SMB loans and closed-end home equity loans.

b. Monetization rates can fluctuate with asset mix, supply-demand, and signing timing, and the goal is to maximize variable profit. See Q&A for demand color.

5. Financial Services

a. Revenue was about $525mn (+25% YoY), with net interest income of $281mn (+38%, driven mainly by deposit growth) and noninterest income of $244mn (+13%).

b. Products totaled 23.2mn (+44% YoY), with Money and Relay both above 8mn (up 34% and 45%, respectively). Invest exceeded 4mn (+37%).

c. Credit cards were about 656k (+67% YoY), crypto users exceeded 500k, and SoFi Plus was around 275k.

d. Interchange revenue hit a record $52mn (+77% YoY), corresponding to annualized spend of $31bn across Money and credit cards. Brokerage commissions were $18mn (+47% YoY), and credit card revenue nearly doubled QoQ.

e. Annualized revenue per product was about $90 (vs. $88 last quarter), with credit cards and crypto contributing the most uplift. SoFi Plus delivers $120 recurring revenue per product per year, about 83% of subscribers are existing members, and 25% add another product post-subscription.

6. Technology Platform

a. Revenue was $87mn (-24% YoY) due to a large client migration. Excluding that client, revenue grew +18% YoY, and nine new clients were added this year.

b. Management remains confident in a return to a steady-state 25% growth rate.

7. Long-term goals

a. Eight businesses now each generate over $100mn annualized revenue (one above $2bn, one above $1bn, three above $300mn, and three above $100mn). Most are still growing 30%+ YoY.

b. Newer initiatives such as crypto, SoFi Plus, SMB lending, enterprise banking, SoFi Pay, and SoFiUSD are not yet in that group of eight.

c. By end-2030, the target is 100mn products (about 4x today). Five businesses are expected to exceed $1bn each, five to exceed $500mn each, and another five to exceed $100mn each.

2.2 Q&A

Q: Target segments and borrower traits for SMB lending, and 2027 prioritization?

A: SMB remains early-stage and is not one of the eight major businesses, and investment will continue with a focus on underwriting and unit economics. Over roughly the past five years, SoFi generated SMB-related revenue via lead flows that began when many members sought PPP loans during COVID, revealing many existing consumers were small business owners.

SMB had been a low-investment 'seed' pilot, and after scale became clear in 2025, SoFi decided to build it as a core offering. It launched in 2026, spanning not only loans but also checking and savings, and management values the re-borrowing dynamic where borrowing fuels growth that leads to repeat borrowing.

Q: Growth assumptions by segment in the 2027 guide?

A: Lending +25%, Technology Platform +25%, and Financial Services +35%, with broadly stable segment margins. The full-year guide was pulled forward as visibility improved, with both revenue and cash revenue at $1.4bn in Q3 providing a durable base.

In late 2025, the company decided to accelerate investments in businesses outside the core eight while covering those costs with revenue and still delivering the profit commitment. The 20% net margin in 2027 is at the low end of the long-term range, EBITDA margin is mid-range, and ROTCE will continue progressing toward the 20%–30% goal.

Q: How do you view member headroom and ARPU ceiling (e.g., 40–50mn potential members and about $30bn TAM)?

A: No confirmation on the TAM figures, and management noted that weighted revenue per product will likely decline over 2026–2030. The company focuses on revenue per product rather than per member, and targets 100mn products with at least two per member, with Avg. revenue per product flat to slightly down and internally projected to decline depending on mix.

By product, Invest revenue-to-AUM lags industry benchmarks, with upside from AUM growth, retirement and trust accounts, and options and margin trading. Credit card monetization is early, avoids costly 0% balance transfers, and effective rates are rising.

Money shows rising Avg. deposits per account, SMB has ample monetization room, and LPB monetization fluctuates within a range, so expectations should be balanced. SoFi Plus is $10 per month today, with potential premium tiers, and loan product revenue per product is broadly stable.

Q: How will you fund 2027 growth internally?

A: Two levers provide an aggregate ~$29bn of RWA capacity, none of which is needed for the 2027 guide. First, existing excess capital: Tier 1 plus Tier 2 is about $9.6bn, which at a 13.5% total capital ratio supports about $71bn of RWAs, roughly $15bn above the current $56bn.

Second, internally generated capital: with Q4 net income around $270mn and $1.28bn in 2027 plus stock comp, that adds roughly another $14bn of capacity at the same ratio. Beyond 2027, rising ROE and internal capital generation will support growth.

Q: If rates fall, is that a tailwind or offset by other factors?

A: Management said the guide is achievable in both high- and low-rate environments, with puts and takes under cuts. In the last hiking cycle, growth was intentionally slowed given 'higher for longer' risks and limited diversification and capital at the time, and in 2026 the company demonstrated it can lift coupons, preserve unit economics, and layer hedges under high rates.

Under cuts, housing should rebound meaningfully, and agency-eligible mortgages are off balance sheet with low capital needs and better margins. However, their contribution to visibility is lower than longer-duration personal and student loans.

Q: Annualizing Q4 by segment and applying 2027 growth yields about $6.7bn vs. the $6.4bn guide. What explains the ~$300mn gap?

A: The difference reflects minor changes in the corporate line. Management provided no further breakdown.

Q: LPB demand backdrop, and status with existing and new partners?

A: Demand remains strong, with Q4 originations pacing ahead of Q3’s $3.8bn, and monetization performing well. Unsecured personal loan demand is growing with both existing and new partners, two new partners were signed this quarter, and several more are in discussions, and last quarter SoFi signed a 3-year $3bn SMB loan deal with BasePoint.

SMB remains a small share but will scale, and the platform has issued its first closed-end second-lien HE loans with demand supporting far greater scale than today. Future growth will come from continued momentum in personal loans and rising demand for other asset types such as second-lien mortgages.

Q: Within the Technology Platform’s 25% growth for 2027, how much is third-party vs. SoFi internal products?

A: No split provided, as the business is not managed by that dimension. The platform was built for SoFi’s most complex internal needs and happens to fit market demand, and four growth pillars are card issuing processing, cloud-native core, payment orchestration, and risk and anti-fraud.

The cloud core supports new member onboarding, enterprise banking, and crypto buy/hold/sell. As a regulated bank, internal SoFi businesses transact with the platform on an arm’s-length basis, so internal revenue effectively replaces costs that would otherwise be paid externally.

Management said the company is the only national bank issuing a payments stablecoin and tokenized deposits, and is moving faster than other fintechs and large banks across the AI and blockchain cycles.

Q: Personal loan originations are growing faster than the industry; can that magnitude be maintained in 2027?

A: More growth in 2027 will come from home loans and student loan refis, and personal loans will not grow at the same pace. Personal loans still have momentum, but mix will tilt toward home loans and student refis, and allocation follows risk-adjusted returns, with personal loan risk-adjusted ROA above 7.5%, over 100bps higher YoY, driven mainly by a 2.5x rise in origination fees and secondly by loss improvement.

Q: Weighted personal loan coupons have trended down; how do you balance coupons, spreads, and origination fees in pricing?

A: Coupon plus fee yield was broadly flat QoQ, and risk-adjusted returns improved about 100bps YoY. Pricing is not optimized on a single variable, and whereas coupons led before, origination fees have become more important, with over 50% of new personal loans carrying fees, and Q3 origination fee revenue at about $270mn, 2.5x YoY and +36% QoQ, mostly from personal borrowers.

Borrowers who pay fees prepay 600–700bps less than those who do not, extending asset duration, and their loss rates are 40–50bps lower. Combined, that adds about 200–300bps to ROE, so the company is pushing up the share of fee-bearing loans.

Q: How to raise checking deposit mix and lower overall funding costs?

A: Do not view deposit rates through a traditional lens; they replace branches and headcount as acquisition costs, and brokered deposits can actually be cheaper. With no branches and no outbound calling, leading deposit rates are part of the digital acquisition value proposition, and SoFi Money is the spearhead for cross-sell, providing data on members’ mortgages, card balances, and idle cash.

Funding tools also include brokered CDs, warehouse lines, securitization, and whole loan sales, and in volatile rate periods, brokered CDs and wholesale funding play a bigger role. Decisions weigh marginal funding cost on new assets vs. repricing costs of member deposits, and long term, 80%–90% of funding is expected to be deposits.

Management added they do not see AI (rebranded internally as SI, super intelligence) as a threat to deposits but as a tailwind, and acquisition costs did not rise even as marketing spend grew linearly, with awareness the main bottleneck for non-capital businesses. When AI agents compare rates, hidden fees, and thresholds, greater transparency favors SoFi, which has the highest lifetime value.

Q: SoFi Plus growth slowed this quarter; is the target still 1mn members by end-2026?

A: The target remains 1mn by end-2027, Q3 issues were fixed, and Q4 is re-accelerating. Adjustments were made in Q3 to ensure pricing and enrollment elements were in place, which weighed on growth, and management said the target will be achieved even if not on the original timetable, as millions of the 17.1mn members should subscribe.

The product will iterate monthly, modeled on Amazon Prime with ongoing enhancements such as higher deposit rates, higher transfer matches, higher card cashbacks with lower APRs, and better mortgages, and subscription cancellation is free (vs. $9.99 charged by peers), and the acquired Composer AI strategies are unlimited. About 85% of subscribers are existing members, and 25% add another product post-subscription.

Q: How will you balance LPB vs. on-balance sheet loans?

A: No preset split, with dynamic allocation based on risk-adjusted returns and durability of revenue, and both will grow. With ample capital, SoFi can retain more loans to maximize long-term shareholder value, while strong LPB partner demand and expanding asset types make LPB attractive as a capital-efficient business, and the mix depends on capital opportunities, market conditions, and partner needs.

Q: With more banks and fintechs going multi-product, how does SoFi sustain its moat?

A: The moat is superior unit economics per product and higher products per member, which together maximize lifetime value. The company does not pursue the lowest operating cost but maximizes unit economics, and each product must underwrite to long-term variable profit even if it loses money early on, as credit cards once did, and the long-term unit economics remain compelling.

With the highest lifetime value, SoFi can offer better prices and services like Walmart and Amazon, driving a 'financial services productivity flywheel'. It does not matter which product a member adopts first because each must stand alone economically, and examples include refinancing federal student loans from 8% to 6% or consolidating 25% credit card debt into a 12% personal loan while still earning attractive returns.

Q: Can fee-based revenue reach 50% and ROTCE 20%–30% by 2030?

A: No direct confirmation on the 2030 numbers, but the path is to lift net margin above 30% and raise the revenue-to-equity ratio. Current ROTCE is 10% and equals net margin multiplied by the revenue-to-equity ratio, and in 2027, net margin is 20% with incremental net margin at 32%, and incremental margin should trend toward the steady-state long-term margin.

The revenue-to-equity ratio is about 57% now, vs. 55% last quarter and 50% the prior quarter, rising with higher fee-based mix and Technology Platform revenue, which is non-capital intensive. As fee-based revenue moves from 42% to above 50%, the revenue-to-equity ratio will approach 1, consistent with a 20%–30% return.

<End of text>

Risk disclosures and statements:Dolphin Research Disclaimer and General Disclosures