Analyst: Bear (PaulWBryant) Date: 2026-03-30 Quarter: Q4 FY25 (Dec-2025) Price: ~17.80(Mar25, 2026)|MarketCap: 24.0B TTM Revenue: $3.613B (GAAP) | TTM Adj Net Revenue: 3.591BP/TTMRevenue: 6.7x|P/FY26ERevenue( 4.655B): ~5.2x FY26E P/E: ~30x (0.60adjEPSguide)|P/TBV: 2.5x(7.01 TBV/share) Shares: 1,346M diluted (+17% YoY)
I haven't written about SoFi before. Let me be clear: I came into this with skepticism. SoFi is a bank holding company, not a SaaS company. Banks typically don't compound at 30%+ revenue growth for extended periods. The bear case on SoFi has always been that it's a rate-sensitive personal lender dressed up in fintech clothing, and eventually credit losses or NIM compression would reveal the fragility underneath.
Going into Q4 FY25, given Q3's trajectory, my expectations were:
| Metric | Prior Belief | Rationale |
|---|---|---|
| Adj Net Revenue | ~$960-980M | Residual of $3.54B FY25 guide minus Q1-Q3 actuals |
| EBITDA Margin | ~29-30% | Continuation of Q3's 28.8%, modest expansion |
| Members Added | ~800-900K | Strong Q3 pace of ~897K |
| Financial Services Revenue | ~$430-440M | Growing 60-70%+ YoY trajectory |
| FY26 Revenue Guide | $4.3-4.5B | ~25% growth — typical conservative step-down |
| Credit Quality | Stable-to-slightly-worse | Macro pressure on consumer loans |
| Tech Platform | Stable accounts | Large client headwind already visible |
| Metric | Expected | Actual | Delta | Assessment |
|---|---|---|---|---|
| Adj Net Revenue | ~$960-980M | $1,012.8M | +$33-53M / +3-5% beat | Better. First $1B quarter. |
| GAAP Revenue | — | $1,025.1M (+39.6% YoY) | — | Strong YoY re-acceleration |
| EBITDA | ~$280-295M (~30%) | $317.6M (31.4%) | +$23-38M | Better. ATH margin. |
| Gross Margin | ~82-83% | 84.2% | +1-2pp | Better. Near ATH (84.5% Q1 FY24). |
| Operating Margin | ~14-16% | 18.1% | +2-4pp | Significantly better. +9.9pp YoY. |
| Members Added | ~800-900K | 1,009K | +100-200K | Better. First >1M quarter. |
| Products Added | ~1.2-1.4M | 1,615K | +200-400K | Better. Cross-buy 40% (record). |
| Financial Services | ~$430-440M | $456.7M (+78% YoY) | +$17-27M | Better. Approaching Lending parity. |
| FY26 Revenue Guide | $4.3-4.5B | $4.655B (+30%) | +$155-355M | Higher than expected. Ambitious. |
| FY26 EPS Guide | — | $0.60 | — | First EPS guide — implies 54% growth |
| Credit (PL charge-off) | 2.9-3.0% | 2.80% | Better | Down 50bps YoY. Up 20bps QoQ — mixed. |
| Tech Platform Accounts | ~150M | 128.5M (-23% YoY) | Worse | Large client fully departed. Real headwind. |
Verdict: The quarter exceeded my expectations on almost every metric. The magnitude of beat — particularly the $1B+ revenue milestone, 31.4% EBITDA margin, and 1M+ member adds — was genuinely impressive. I could be wrong, but this looks like a clean inflection quarter.
| Q1_FY24 | Q2_FY24 | Q3_FY24 | Q4_FY24 | Q1_FY25 | Q2_FY25 | Q3_FY25 | Q4_FY25 | |
|---|---|---|---|---|---|---|---|---|
| Revenue ($M) | 645.0 | 598.6 | 697.1 | 734.1 | 771.8 | 854.9 | 961.6 | 1,025.1 |
| YoY % | +36.6% | +20.2% | +29.8% | +19.3% | +19.7% | +42.8% | +37.9% | +39.6% |
| QoQ % | +20.1% | -7.2% | +16.5% | +5.3% | +10.7% | +10.8% | +12.5% | +6.6% |
| Adj Net Rev ($M) | 580.6 | 597.0 | 689.4 | 739.1 | 770.7 | 858.2 | 949.6 | 1,012.8 |
| Adj YoY % | +9.5% | +12.6% | +30.0% | +24.4% | +32.7% | +43.7% | +37.7% | +37.0% |
| Gross Margin | 84.5% | 81.7% | 82.3% | 82.5% | 82.4% | 82.4% | 83.2% | 84.2% |
| EBITDA ($M) | 144.4 | 137.9 | 186.2 | 198.0 | 210.3 | 249.1 | 276.9 | 317.6 |
| EBITDA Margin | 22.4% | 23.0% | 26.7% | 26.8% | 27.2% | 29.1% | 28.8% | 31.4% |
| Net Income ($M) | 88.0 | 17.4 | 60.7 | 332.5* | 71.1 | 97.3 | 139.4 | 173.5 |
| EPS (adj) | — | — | — | — | 0.06 | 0.08 | 0.11 | 0.13 |
| Shares (M) | 1,042.5 | 1,065.2 | 1,104.5 | 1,151.0 | 1,185.5 | 1,182.9 | 1,291.0 | 1,346.1 |
*Q4 FY24 GAAP NI includes $271.5M one-time DTA recognition. Adj NI was $61.0M.
| Q1_FY25 | Q2_FY25 | Q3_FY25 | Q4_FY25 | |
|---|---|---|---|---|
| Financial Services | 303.1 | 362.5 | 419.6 | 456.7 |
| Lending | 413.4 | 443.5 | 493.4 | 498.7 |
| Tech Platform | 103.4 | 109.8 | 114.6 | 122.4 |
| Total | 771.8 | 854.9 | 961.6 | 1,025.1 |
The numbers have to match the theory. And in Q4 FY25, they did. Let me work through what matters and what worries me.
This is the most important story in the quarter, and Atlas flagged it correctly. Financial Services revenue at 456.7M(+78498.7M, +19.4% adj) as the largest segment. Full-year numbers make it even clearer:
At current trajectories, Financial Services surpasses Lending by Q1 or Q2 FY26. In other words, SoFi is becoming a financial services platform that also does lending, not a lender that also does financial services. This matters for valuation — FS revenue carries higher margins (contribution margin 50.5% vs Lending 54.5%) and lower capital intensity.
The LPB is the secret weapon here. $193.7M in Q4 revenue, annualizing at $775M, up 2.9x YoY. This is capital-light fee revenue with no credit risk to SoFi. It's essentially a marketplace for personal loan originations. If LPB scales to $1B+ annualized in FY26 (plausible at Q4 exit rate), it fundamentally changes the capital efficiency of the lending operation.
40% of new products opened by existing members. Up 7 percentage points year over year. Products per member ticked to 1.48 from 1.46. Revenue per product reached $104 annualized, up 29% YoY.
In other words, SoFi isn't just acquiring members — it's monetizing them at an accelerating rate. This is the Financial Services Productivity Loop (FSPL) that Noto talks about, and the numbers now back the theory. A member who has SoFi Money, Invest, and a personal loan is significantly stickier than one with just a savings account.
Nine consecutive quarters of GAAP profitability. EBITDA margin hit 31.4% — above the 30% long-term target management set. Operating margin at 18.1% (+9.9pp YoY). Incremental EBITDA margin of 44% in Q4. The company generated $481M in GAAP net income for FY25.
The Rule of 40 score at 68 is exceptional. For context: 37% adj net revenue growth + 31.4% EBITDA margin = 68.4. Most high-growth SaaS companies aspire to hit 40. SoFi is at 68 as a bank.
Tangible book value $8.9B, $7.01/share (+57% YoY). Capital ratio 22.9% vs 10.5% regulatory minimum — that's a 1,200bp buffer. Cash of $4.9B. Deposits at $37.5B growing $4.6B in a single quarter. Warehouse lines fully paid down.
The price for this fortress was significant dilution. Shares outstanding reached 1,346M, up 195M or 17% YoY. The $1.5B equity offering in Q4 was priced at $27.50 — shares now trade at 17.80, meaninganyonewhoboughttheofferingisunderwaterby35110M annualized savings), and the math checks out. But dilution is dilution. The share count has grown from 963M at Q4 FY23 to 1,346M now — 40% more shares in two years. The EPS CAGR guidance of 38-42% (2025-2028) incorporates this dilution level, but any future capital raises would compress those returns.
Personal loan charge-offs at 2.80% are down 50bps YoY but up 20bps sequentially. Management frames the QoQ increase as mix-driven (more LPB loans in portfolio seasoning) rather than credit deterioration, and the cumulative loss data supports this — Q4 2022 to Q1 2025 vintages show 4.55% cumulative losses vs 6.27% for the 2017 vintage at the same point, and the gap has widened for 6 consecutive quarters.
I don't invest based on hope, so let me state the risk plainly: personal loan charge-offs in a benign environment running near 3% means in a recession they could run 5-7%. SoFi's balance sheet has $37.5B in deposits funding a loan book of comparable size. A credit event would hit both provisions and NII simultaneously. The 22.9% capital ratio provides substantial buffer, but it's never been tested in a real downturn.
The scuttlebutt on this is concerning. SoFi is moving premium features behind a 10/monthsubscriptionthatwaspreviouslyfreeviadirectdeposit.Themathdoesn′tworkformostmembers : the0.2550/year on $20K vs $120/year subscription cost. Doctor of Credit has already flagged the unfavorable economics, and the consumer finance community is circulating "backup bank" advice.
This is a classic "improve monetization at the cost of growth" trade-off. If member adds decelerate meaningfully in Q1/Q2 FY26, we'll know the paywall was a mistake. If members stay because the product suite is sticky enough, it's a brilliant revenue lever. I'd watch products per member very closely — if it drops below 1.45, the cross-buy flywheel is being disrupted.
Nubank received conditional OCC approval in January 2026. Revolut applied in March 2026. These are serious competitors with massive customer bases (Nubank: 100M+ globally). SoFi's bank charter advantage — which has been a genuine differentiator in cost of funds and product flexibility — has a shrinking exclusivity window.
SoFi's response is to build new moats (crypto, stablecoin infrastructure, AI) before the old one closes. The crypto stablecoin play (SoFiUSD) is genuinely differentiated — first from a nationally chartered bank — but it's also genuinely unproven. Revenue contribution from crypto is effectively zero right now.
128.5M accounts vs 167.7M in Q4 FY24 — a 23% decline. Management guides Tech Platform at ~20% growth excluding the departed client, but they won't provide a clean baseline, which makes it impossible to verify. The deflection on Goldman's question about the termination fee characterization is the most uncomfortable moment in the call.
Tech Platform is the segment that should be valued on SaaS-like multiples. Losing a major client and refusing to provide clean comparable numbers undermines that narrative.
"Couldn't be more excited" (3x). "Just getting started" (3x). "Unprecedented" (3x). "Record" (15+ times). "Fresh horses we ride" (closing). I'm not a tone police officer, but I've learned to get nervous when management uses this much superlative language. The best quarters are described plainly. This felt like a sales pitch.
That said, Noto backed words with action — the $1M insider buy on March 2 at $17.50-18.20 is meaningful. He didn't have to do that.
January: ~5% workforce reduction (collections automation + offshoring). February: 65 mortgage ops jobs. Both framed as minor individually, but together they represent more active restructuring than disclosed. The offshoring disclosure is particularly notable given existing customer service complaints. I want to see this translate to margin improvement, not just cost-cutting that degrades service quality.
| Metric | Value | Context |
|---|---|---|
| Price | ~$17.80 | Down 31% YTD, 45% from Nov-2025 peak |
| Market Cap | ~$24.0B | |
| P/TTM Revenue | ~6.7x | On $3.613B |
| P/FY26E Revenue | ~5.2x | On $4.655B guide |
| FY26E P/E | ~30x | On $0.60 adj EPS guide |
| P/TBV | ~2.5x | On $7.01 TBV/share |
| EV/TTM Revenue | ~5.8x | ~$20.9B EV (net cash $3.1B) |
| Revenue-PEG | ~0.17 | P/S 5.2 / growth 30% = cheap by my heuristic |
Applying my revenue-PEG heuristic: if P/S (5.2x) is meaningfully below the growth rate (30%), the valuation is reasonable. That ratio of 0.17 suggests the stock is not expensive on a growth-adjusted basis. For comparison, Nu Holdings trades at ~10x revenue growing ~50%, giving a revenue-PEG of ~0.20. SoFi is cheaper on this measure.
The 30x FY26 P/E is not cheap in absolute terms, but if the 38-42% EPS CAGR through FY28 delivers, you're looking at ~$1.08-1.18 in FY28 EPS, putting the FY28 P/E at ~15x on today's price. Is this really a $24B company? Given $3.6B in trailing revenue growing 30%+, first $1B EBITDA year, first bank-issued stablecoin, and a 13.7M member flywheel — yes, I think $24B is defensible. But "defensible" is different from "compelling bargain."
The stock trades below the $27.50 offering price by 35%. Noto bought at 17.50 − 18.20.AtlascalledP/TBV2.0 − 2.2x( 14-15) as an attractive entry. I think the current price is fair value, perhaps slightly below, but not the screaming buy it would be at Atlas's level.
Agree:
Disagree / Add nuance:
The surfaced learning about seasonal QoQ delta is directly applicable here. Q4 QoQ growth decelerated to +6.6% from +12.5% in Q3 — looks concerning at first glance. But the same-quarter comparison tells a different story:
| Quarter | FY23 QoQ | FY24 QoQ | FY25 QoQ | FY24-to-25 Delta |
|---|---|---|---|---|
| Q1 | +11.4% | +20.1% | +10.7% | -9.4pp |
| Q2 | +5.5% | -7.2% | +10.8% | +18.0pp |
| Q3 | +7.9% | +16.5% | +12.5% | -4.0pp |
| Q4 | +14.6% | +5.3% | +6.6% | +1.3pp |
Q4 is typically a softer sequential quarter (year-end effects). The same-quarter QoQ delta improved +1.3pp — not dramatic, but the direction is positive. The real improvement was in Q2, where the FY24-to-25 delta was +18pp, reflecting the elimination of the Q2 FY24 fair-value mark-to-market distortion. The sequential story is stable-to-improving when viewed through same-quarter comparison. No alarm here.
| Metric | FY25 Actual | FY26 Guide | Implied Growth | Achievable? |
|---|---|---|---|---|
| Adj Net Revenue | $3.591B | $4.655B | +29.6% | Plausible — Q4 exit rate annualizes to ~4.05B, soneed 600M acceleration |
| Adj EBITDA | $1.054B | $1.600B | +51.8% | Ambitious — requires 34% margin vs 29% FY25 |
| Adj Net Income | $481M | $825M | +71.6% | Very ambitious — requires 18% margin vs 13.4% FY25 |
| Adj EPS | $0.39 | $0.60 | +53.8% | Consistent with NI growth, assumes minimal further dilution |
The revenue guide at 4.655Brequiresquarterlysequentialaddsaveraging 266M per quarter above the Q4 FY25 run rate. Is that realistic? Q4 adj net revenue was $1,012.8M. Q1 guide is 1, 040M(+2.71.2B+ quarterly revenue by Q4 FY26. Given FS segment growing 40%+ and LPB scaling, this is plausible but tight.
The EBITDA guide at $1.6B / 34% margin is the more ambitious piece. FY25 EBITDA margin was 29%. Getting to 34% requires significant operating leverage. If they're cutting costs via layoffs and offshoring while growing revenue 30%, the math can work — but it means the restructuring isn't cosmetic, it's essential to hitting the guide.
The medium-term >=30% revenue CAGR (2025-2028) implies FY28 revenue of ~$7.9B. From $3.6B to $7.9B in three years as a bank. Shows what I know, but I'd want to see two more quarters of execution before taking that at face value.
Thesis: Intact. The numbers were excellent, the mix shift is structurally positive, and the valuation at 30x FY26 P/E with 30%+ revenue growth is not stretched.
Conviction: 3.5/5. The business execution is strong, but (a) this is a bank with credit risk I can't fully model, (b) the SoFi Plus paywall creates near-term member growth uncertainty, (c) the charter moat is narrowing, and (d) I haven't seen this management team navigate a downturn. I don't let enthusiasm get ahead of conviction.
Position: No position currently. The stock at ~$17.80 is fair value — not expensive, not a screaming buy. I'd consider starting a small position (2-3%) on a pullback toward $15 (P/TBV ~2.1x) where the margin of safety is more comfortable for a bank holding company with this level of execution. At current prices, I'd watch rather than buy.
What I need to see next:
| Commitment | Due | My Expectation |
|---|---|---|
| Q1 FY26 adj net rev ~$1.04B | Q1 FY26 ER (~Apr 2026) | Should meet — conservative step |
| FY26 adj net rev ~$4.655B | FY26 end | Achievable if FS maintains 40%+ |
| FY26 adj EPS ~$0.60 | FY26 end | More challenging — needs margin expansion |
| Member growth >=30% FY26 | FY26 end | At risk from SoFi Plus paywall |
| SoFi Invest full profitability | FY26 end | Plausible given 2.2x brokerage rev growth |
| NIM above 5% | Ongoing | Likely if Fed holds; risk if aggressive cuts |
| Revenue CAGR >=30% (2025-2028) | FY28 | Ambitious but not implausible |
Bear