SoFi's 47.2% drawdown prices a credit-cycle casualty the Q1 2026 print disproves

SoFi Technologies (SOFI $17.07 -3.23%) closed July 20, 2026 at $17.01, down 35.0% year-to-date and off 47.2% from its 52-week high of $32.21, with 14.7% of shares sold short and a 37.8x trailing P/E on $608.5M of TTM revenue. The drawdown sits against a quarter in which SoFi Technologies added a record 1,055,000 members, its largest quarterly add on file, bringing total members to 14.7 million, up 35% year-over-year, while the Q1 2026 earnings release booked $693.0M of net interest income and a 5.94% net interest margin, 22 basis points above the prior quarter.

The thesis: that 47.2% drawdown from high is the price of a single demand-versus-supply decision SoFi has not yet made: how much of its $8.3B record personal-loan origination quarter to hold on the balance sheet versus sell down to loan-platform buyers, and the evidence that decision is already tilting toward hold is more concrete than the short base implies. Net interest margin expanded 22 basis points quarter-over-quarter despite a 63 basis-point decline in average asset yields, because cost of funds fell 48 basis points and management quantified the deposit-versus-warehouse remixing as $621.8M of annualized interest expense savings: argues the trade is being made, and made quickly enough that the multiple, not the fundamentals, is what has to give.

Key takeaways

• The 37.8x trailing P/E on $21.8B of market cap and a -35.0% YTD drawdown sit on +23.1% trailing revenue growth, a steeper revenue print than Visa Inc. (V $353.42 -0.67%) at +12.3%, Mastercard Inc (MA $531.98 -1.17%) at +16.4%, or American Express Co (AXP $348.74 -0.58%) at +13.4% on the peer table.
• The 47.2% peak-to-trough drop and 14.7% short interest are pricing SoFi as a credit-cycle casualty while the Q1 2026 release shows a 22 bps NIM expansion to 5.94%, a record 1,055,000 net member adds, and management framing on the Q1 2026 call that capital headroom, not demand, is the gating variable.
• Tangible book value of $7.21 per share, up 57% year-over-year, and SoFi Bank CET1 of 15.3% set a structural floor more than 8 percentage points above the 7.0% required minimum, and management is reinvesting into the loan book rather than returning capital because the Q1 2026 10-Q confirms no funds free of restrictions are available for dividends.
• Full-year 2026 guidance is at least 30% member growth, adjusted net revenue of approximately $4.655B, adjusted EBITDA margin of approximately 34%, and adjusted EPS of approximately $0.60 per the Q1 2026 release, a cadence the Q1 print already outpaces on members and products.
• The bull case wins on trajectory but the bear case wins on margin of error, with the asymmetric ceiling bounded by the multiple peers are not being asked to pay and the asymmetric floor bounded by the drawdown the table has already absorbed.

The market is treating SoFi as a credit story, not the consumer platform it is becoming

Price performance vs peers (Credit Services and cross-industry comps)

A 47.2% drawdown from the 52-week high, a 14.7% short interest that is an order of magnitude above Visa, Mastercard, and American Express, and a 35.0% year-to-date decline: those are the price tags the market is putting on a bank that, on the same data set, just posted record originations across every lending product it sells. SoFi's 52-week change of -19.6% is steeper than Mastercard's -1.3% and Visa's +2.7%, and the YTD return is more than 20 percentage points worse than Capital One Financial Corp (COF $199.96 -0.70%), the most direct credit-services comparable, which itself is at -14.4% the performance table. The market is pricing this name like the credit cycle is breaking SoFi, not like a digital consumer platform is widening its funnel.

Two pieces of evidence cut against that read. First, the growth side: the Q1 2026 earnings release added 1,055,000 members in a single quarter, the largest add on file, taking total members to 14.7 million from 10.9 million a year earlier, while total products rose to nearly 22.2 million from 15.9 million: +39% year-over-year. Revenue growth in the same table is +23.1% YoY, against the credit-services peer set of Visa Inc. at +12.3%, Mastercard Inc at +16.4%, American Express Co at +13.4%, and Capital One Financial Corp at +36.6% (where Capital One's print is acquisition-driven, not organic). On the only organic-growth metric in the table, SoFi outruns the network operators.

Second, the funding side, which is where the bear case actually has to land. The Q1 2026 release reported a 48 basis-point decline in cost of funds that more than offset a 63 basis-point decline in average asset yields, leaving net interest margin at 5.94%, up 22 basis points quarter-over-quarter, and quantified the deposit-versus-warehouse remixing as $621.8 million of annualized interest expense savings. The short base is a bet that SoFi cannot deploy the resulting capital into loans at acceptable credit returns; the same release booked $8.3 billion of personal-loan originations (a record), $2.6 billion in student-loan originations (+2.2x year-over-year), and $1.2 billion in home-loan originations (+2.4x). SoFi Bank's CET1 ratio of 15.3% sits 8.3 percentage points above the 7.0% required minimum, per the Q1 2026 10-Q, and there is no sign of capital rationing in the origination lines.

The cross-industry comps sharpen the gap. OppFi Inc. (OPFI $9.24 -2.33%) trades at a 4.6x P/E on a $794.2M market cap with the table showing no revenue growth figure, Yiren Digital Ltd. (YRD $1.25 +0.81%) sits at $109.4M of market cap and -66.2% YTD, and PicS N.V. (PICS $11.91 -3.64%) carries $1.5B of market cap with no P/E in the table. SoFi's 37.8x P/E is a premium, but it is a premium attached to a 23.1% top line against an industry where the largest names are in the low teens. The 14.7% short interest is closer to the small-cap subprime lenders than to the networks SoFi is increasingly being compared to on growth.

The setup, then, is a short thesis built on credit-cycle and capital-deployment fears, trading against a quarter in which demand records landed on every lending product, cost of funds moved in the right direction, and capital headroom was wider than at any time in the company's history. The market's framing is a bank; the disclosure profile is a platform widening a funnel.

What management teams are saying

Management tone splits cleanly down the credit-services peer set. The networks describe volume breadth and pricing tailwinds as the quarter's story; the card issuers describe integration costs and fee-rebuild work as the price of unlocking the next leg of growth. SoFi Technologies sits in neither camp cleanly, and that is the signal.

Ryan McInerney, CEO of Visa Inc., on the Q2 FY2026 call described the quarter as "the strongest net revenue growth since 2022," driven by payments volume up 9% in constant dollars to $3.7 trillion and cross-border volume up 11%. Chris Suh, CFO of Visa Inc., attributed service-revenue growth of 13% versus volume growth of 8% to "pricing and card benefits", the cleanest pricing tailwind in several quarters, while flagging that some of the 24% commercial-and-money-movement revenue growth was one-time "performance adjustments and deal timing" and should not repeat. The tone is "drivers remain strong and relatively consistent with q1." U.S. debit grew 7%, credit grew 10%, and management explicitly said it did "not see signs of the lower spend consumer weakening."

Michael Miebach, CEO of Mastercard Inc, on the Q1 2026 call said "consumer spending is healthy" and that the company "outperformed our own expectations." Sachin Mehra, CFO of Mastercard Inc, confirmed pricing was a positive contributor on every line called out: cross-border assessments up 18% versus 13% volume growth "driven primarily by pricing in international markets", and said "all of our planned pricing is already contemplated in the guidance." Switch transactions eased from 10% in Q4 2025 to 9% in Q1 2026, attributed to three transient factors: the GCC/Israel conflict, the completion of the Capital One debit migration, and Ramadan/Easter timing. On the Bernstein investor event on 2026-05-28, Mehra added that growth "has continued into May, the first two weeks as well, where we saw, you know, stable to slightly better consumer spending."

The card issuers frame the same period as a build year. Stephen Squeri, CEO of American Express Co, on the Q1 2026 earnings call said card-member spending "grew 10% on a reported basis… the highest quarterly growth in 3 years," driven by T&E and goods and services. The tone then pivots: management chose to reinvest the upside into marketing and technology rather than raise the 9-10% revenue and $17.30-$17.90 EPS guidance range. Squeri called the Platinum refresh a spend accelerant: "about 1/4 of the overall U.S. consumer Platinum portfolio has been built for the higher annual fee, and we have seen no change to our very high retention rates", and framed commercial product launches, including the Graphite Business Cash Unlimited card, as 2027 tailwinds rather than 2026 contributors.

Rich Fairbank, CEO of Capital One Financial Corp, on the Q1 2026 call described purchase volume growth of 40% year-over-year, but only 8% excluding the Discover acquisition. Ending card loans were up 69% year-over-year but only 3.9% on the same basis. Auto originations were up 21% year-over-year, and consumer banking loan balances were up about 10%: the underlying organic book is growing, but the headline number is the Discover print. Fairbank framed the Discover integration as a multi-quarter exercise: "we expect to have fully transitioned new originations by the end of Q3," with the back book converting through Q1 2027 and full $2.5 billion of synergies "by the time we complete the integration in the middle of 27." Management is leaning on Brex as an enablement play rather than an EPS contributor in the near term.

SoFi Technologies' management sounds like neither. Anthony Noto, CEO of SoFi Technologies, on the Q1 2026 call said the gating question is "how much" personal-loan volume to hold versus sell to loan-platform buyers, not whether demand exists: "we had record originations in personal loans and student loan refinancing… our home loans business more than doubled year-over-year." On capital, Noto said the bank has "a lot of capital on the balance sheet… adequate capital would be nuts not to be able to meet that demand and hold some of it on our balance sheet and continue to generate increasing net interest income." That is a demand-side tone wearing a capital-deployment answer: closer in posture to Visa Inc. and Mastercard Inc describing healthy volume than to Capital One Financial Corp describing integration drag, and it does not match the credit-constructive-but-fee-rebuilding tone at American Express Co.

The cross-industry comps diverge further. Eduardo Chedid, CEO of PicS N.V., on the Q1 2026 release said the company "exceeded guidance across key indicators," and management set Q2 2026 guidance for +13% sequential managerial revenue and +45% sequential adjusted net income off a 1Q26 base that itself beat prior guidance on every line. The tone is "profitable growth" with a deliberate higher-risk, higher-spread loan book. Todd Schwartz, CEO of OppFi Inc., on the Q1 2026 call was more measured: new-customer originations up 8% year-over-year but total retained net originations down 10%, with receivables still up 9.4% on a higher starting base, and net charge-offs at 55.5% of average receivables characterized as "more of a normalization" against FY2024-FY2023 vintages. The BNC acquisition, expected to close in Q4 2026, is "at least 25% accreted to adjusted EPS in the first year post-closing."

Ning Tang, CEO of Yiren Digital Ltd., on the Q1 2026 call framed demand as recovering from a regulatory shock: "our credit solution business continued to recover as industry credit conditions improved following a year of challenging regulatory tightening and credit normalization." Repeat borrowing reached a record 78% of loan volume, and the insurance segment returned to growth on both sequential and year-over-year basis. William Hui, CFO of Yiren Digital Ltd., said cost of capital has decreased 93 basis points since the new regulation took effect in October 2025 and characterized the credit cycle as having peaked in October with "signs of recovery" since. The tone is bottoming, not breaking: closer to SoFi's "supply of capital is plentiful" framing than to the bear read embedded in SoFi's 14.7% short interest.

The cleanest contrast on demand tone sits between SoFi Technologies and Capital One Financial Corp: Fairbank discusses how much Discover volume to absorb through a multi-quarter conversion, while Noto discusses how much of a record personal-loan quarter to hold on balance sheet versus sell down to the loan platform. American Express Co reinvested its beat; Visa Inc. and Mastercard Inc attributed theirs to pricing and breadth. SoFi's beat was on volume, and management's stated plan is to put more of it on balance sheet, a posture none of the peers are taking, because none of the peers sit at SoFi's combination of deposit-funded funding cost (down 48 basis points year-over-year) and 15.3% CET1 capital headroom.

The pricing commentary across the peers lines up where the bear case on SoFi does not. Visa Inc., Mastercard Inc, and American Express Co are all describing pricing as a positive contributor to revenue growth this quarter. SoFi Technologies is describing pricing in the opposite direction: a 63 basis-point decline in average asset yields. The difference is mechanical, SoFi lends, the others route, but it explains why the market is conflating SoFi with the credit cycle rather than with the network operators. The SoFi story has to run on volume plus NIM defense, and Q1 2026's NIM expansion came from cost-of-funds improvement, not from loan yields. That is the line the bear case is leaning on, and the data on originations and member growth is what the bull case has to point to.

The valuation premium is paying for a balance-sheet platform, not a credit bet

Valuation vs peers (Credit Services and cross-industry comps)

Performance vs peers (Credit Services and cross-industry comps)

SoFi Technologies, Inc. trades at 37.8x trailing earnings on a $21.8B market cap and $608.5M of trailing-twelve-month revenue, against a peer set that ranges from Mastercard Inc at 31.7x to American Express Co at 22.0x and a 4.6x print at OppFi Inc.. The multiple gap is the question; the comparison data answers it. SoFi is the only name in the table where the trailing revenue figure sits below $1B while the P/E sits above 30x, and the only name where revenue growth of 23.1% year-over-year runs alongside a market cap more than two-and-a-half times OppFi Inc.'s, PicS N.V.'s, and Yiren Digital Ltd.'s combined. The premium is paying for a balance-sheet platform that turns deposits into loans, not for a routing business that turns transactions into fees.

Mastercard Inc at 31.7x carries $33.9B of trailing revenue and 16.4% revenue growth; American Express Co at 22.0x carries $42.2B and 13.4% growth; Visa Inc. at a market cap of $646.8B carries $43.0B of trailing revenue and 12.3% revenue growth. The network operators are priced on revenue scale and steady low-teens growth. Capital One Financial Corp at 52.8x and 36.6% revenue growth is the only name in the table with a richer top-line print than SoFi, and that print is the Discover acquisition flowing through, not organic growth; Fairbank's own commentary on the Q1 2026 call put ex-Discover purchase volume growth at 8%. On the only organic-growth metric in the table, SoFi at +23.1% leads the credit-services peers and runs close to the networks' pricing-driven prints.

The comparison tightens when management's own guidance is read alongside the table. the Q1 2026 release set full-year 2026 guidance for adjusted net revenue of approximately $4.655B, implying approximately 30% annual revenue growth, with adjusted EBITDA of approximately $1.6B at a margin of approximately 34% and adjusted EPS of approximately 60 cents per share. Management expects members to grow at least 30% year-over-year, an explicit trajectory above the +23.1% in the table. The Q2 2026 cadence management laid out: approximately 30% adjusted net revenue growth and approximately 30% adjusted EBITDA margin the Q1 2026 release. SoFi is not guiding to a deceleration that would justify the short base; it is guiding to acceleration from the trailing print, with member additions already running above the annual target in Q1 alone.

The small-cap comps sharpen the gap rather than narrow it. OppFi Inc. carries a 4.6x trailing P/E on $794.2M of market cap, with the table showing no revenue growth figure at all, and management on the Q1 2026 call framed total retained net originations down 10% year-over-year even as new-customer originations rose 8%. Yiren Digital Ltd. at $109.4M of market cap is -66.2% year-to-date and -81.1% off the 52-week high, with management on the Q1 2026 call reporting loans facilitated down 42% year-over-year at RMB 8.9B and a net loss of RMB 494.7M; the Q1 2026 release framed the cycle as recovering from a regulatory shock, but the table price is a year-to-date decline of 66.2%. PicS N.V. sits at $1.5B of market cap with no P/E or revenue growth on file in the table. None of these names is a pricing comparable to SoFi; the multiple is a category error, and reading SoFi against them rather than against Visa Inc. or Mastercard Inc is what produces the "47.2% off the high" framing in the first place.

The capital-return line in the table tells the same story the multiple does. SoFi has no dividend on file in the table against peers who return material cash. Visa Inc. returned $11.7B in buybacks over the first half of fiscal 2026 per the Q2 FY2026 10-Q and authorized a fresh $20.0B repurchase program in April 2026; Mastercard Inc repurchased $4.035B at $519.67 per share in Q1 2026 per the Q1 2026 10-Q and raised the quarterly dividend 14.5%; American Express Co returned $2,316M in Q1 2026 and raised its quarterly dividend 16% to $0.95 per share per the Q1 2026 10-Q; Capital One Financial Corp returned $3.0B in Q1 2026 ($505M dividends plus $2.5B repurchases) per the Q1 2026 10-Q. SoFi's capital base sits at $10.8B of equity with TBV/share of $7.21, up 57% year-over-year per the Q1 2026 release, and management's stated posture is reinvestment into loans and a new core banking platform rather than return of capital; the Q1 2026 10-Q confirms no funds free of restrictions for dividend payments at the bank or the holdco as of March 31, 2026. The valuation premium is buying growth, not yield, and the table price tag, -35.0% year-to-date, is the market's discount on that bet.

The peer-set read on performance tightens the case. Visa Inc. is +3.0% year-to-date, American Express Co is -4.7%, Mastercard Inc is -4.0%, and Capital One Financial Corp is -14.4% performance vs peers. SoFi's -35.0% year-to-date is more than 20 percentage points below Capital One Financial Corp, the closest credit-services comparable on revenue scale, and is the deepest drawdown in the table aside from Yiren Digital Ltd.'s -66.2%. The 52-week change of -19.6% for SoFi is steeper than Mastercard Inc's -1.3% and Visa Inc.'s +2.7%, and the 14.7% short interest is an order of magnitude above the network operators' sub-2% prints. The market is pricing SoFi as a credit-cycle casualty, and the disclosure profile: record originations across personal, student, and home loans, NIM expansion on cost-of-funds improvement, and explicit guidance for 30% member growth in 2026: argues the trade is being made faster than the table's price reflects.

The balance sheet has room to grow into the guided loan book

SoFi Technologies, Inc. balance sheet — as of Mar 31, 2026 (10-Q filed May 7, 2026)

Capital is plentiful, but the working configuration skews liabilities-heavy by design: deposits and securitized warehouse facilities fund the loan book, equity sits well above regulatory minima, and cash on hand is small relative to total assets because SoFi runs a balance-sheet-led model rather than a fee-led one.

Cash and equivalents stood at $3.4 billion against total assets of $53.7 billion at March 31, 2026, per the Q1 2026 10-Q. That ratio, a low single-digit percent of assets in cash, is consistent with how a chartered bank holds liquidity rather than how a payments network does; the comparable network operators carry a meaningfully larger cash buffer relative to their much smaller balance sheets. Long-term debt of $427.0 million is modest against $10.8 billion of total stockholders' equity and $53.7 billion of assets, leaving equity as the dominant capital cushion rather than debt. The goodwill line of $1.4 billion is the legacy of the Galileo and Technisys acquisitions and represents a small share of tangible capital; tangible book value per share of $7.21, up 57% year-over-year, is the cleaner read on the equity base.

The capital story sits in the regulatory ratios. SoFi Bank's CET1 ratio of 15.3% and Tier 1 leverage of 12.8% per the Q1 2026 10-Q both run multiples above the 7.0% and 4.0% required minimums. SoFi Technologies on a consolidated basis carries CET1 of 21.1% and Tier 1 leverage of 17.7%. Both ratios stepped down from year-end 2025, Bank CET1 from 16.4%, holdco CET1 from 22.8%, reflecting balance-sheet growth rather than capital erosion: the equity base grew by $322.1 million during the quarter per the Q1 2026 release, even as risk-weighted assets expanded to absorb the origination records. Management's framing on the Q1 2026 call was that the question is the balance-sheet-versus-loan-platform split on a given origination, not whether capital exists to absorb it. The 48 basis-point decline in cost of funds and the deposit-versus-warehouse remixing that management quantified as roughly $621.8 million of annualized interest expense savings in the Q1 2026 release was a funding-side improvement that ran ahead of asset growth.

Dividend capacity is explicitly constrained. The Q1 2026 10-Q states that as of March 31, 2026, neither SoFi Bank nor the holding company had funds free of restrictions available for dividend payments, and no common-stock dividend has been declared. That posture is consistent with a growth-stage bank choosing to retain earnings against the guided 30% member-growth target and the loan-platform-versus-balance-sheet deployment choice management is making quarter by quarter. The forward-looking statements in the Q1 2026 release reference settlement of the 2026 convertible notes as a planned item, indicating that debt management, not buybacks or dividends, is the return-of-capital lever on the 2026 calendar.

The cleanest read across the disclosures is that the balance sheet is built for the next leg of originations rather than for capital return. Tangible book compounding, deposit-funded cost of funds, and regulatory headroom that exceeds required minima by multiples together point to a balance sheet that can absorb a larger loan book at acceptable credit returns without forcing a capital action. The bear case has to argue that the marginal loan yields will not justify the marginal capital tied up against them; the Q1 2026 record originations, NIM expansion, and cost-of-funds improvement argue the marginal trade is still clearing.

Three things have to land — and any one of them breaking cleanly sinks the trade

The bull case rests on three load-bearing pieces. First, the loan-versus-loan-platform deployment mix has to keep tilting toward balance-sheet hold. Management's framing on the Q1 2026 call was explicit: "if we're going to put loans on our balance sheet, they obviously require capital and they have credit risk… LPB revenue, on the other hand, doesn't require capital." The 22 basis-point NIM expansion to 5.94% and the 48 basis-point decline in cost of funds quoted in the Q1 2026 release sit on the assumption that SoFi keeps earning the spread on a growing slice of that $8.3B personal-loan quarter. If member growth of at least 30% in 2026 materializes but most of it leaves through the loan platform, the multiple compression runs faster than the funding-cost tailwind can offset it, because fee LPB revenue scales sub-linearly against the 37.8x P/E in the table.

Second, credit has to hold within the corridor implicit in NIM expansion through a down-yield cycle. The Q1 2026 release reported a 63 basis-point decline in average asset yields, meaning SoFi is accepting a thinner gross loan yield to keep volume moving; that is the trade management is making, and it is the trade that breaks if delinquency trends push provisions higher. SoFi Bank's CET1 ratio of 15.3% gives more than eight percentage points of headroom above the 7.0% required minimum per the Q1 2026 10-Q, but headroom is not the same as absorption capacity if credit normalizes against a still-growing $53.7B asset base. The bear case at Capital One Financial Corp, where the Q1 2026 call flagged heightened geopolitical uncertainty pushing the company to greater consideration of downside economic scenarios, is the template: reserve build compresses NII even when originations are records.

Third, the second-half 2026 cadence has to deliver on the "approximately 30%" adjusted net revenue growth and "approximately 30%" adjusted EBITDA margin guides from the Q1 2026 release. Noto's own framing on the Q1 2026 call was that "this is obviously a year in which revenue is going to be disappointing because of a loss to that large customer," with a stated objective of "20% to 25% compounding growth over the years to come." If the loss-to-large-customer gap doesn't close, the 23.1% revenue growth in the performance table becomes the ceiling rather than the floor, and the 37.8x P/E starts to require a multiple of the same customer recovery that the networks at Visa Inc. and Mastercard Inc are not being asked to price in.

The bear case on each of those three pieces is concrete. On deployment mix, the Q1 2026 release called out $3.6B of new LPB commitments in the quarter alongside the balance-sheet build, and management described the platform revenue stream as "cash flow up front" with no retained credit risk; the more LPB capacity grows, the more the marginal loan can leave the balance sheet, and the more the NIM expansion story depends on a shrinking share of total originations. On credit, Yiren Digital Ltd. management on the Q1 2026 call described a regulatory tightening that compressed loan volumes 42% year-over-year even as cost of capital fell 93 basis points, a reminder that funding tailwinds can co-exist with credit-cycle deterioration. SoFi's own 14.7% short interest is the market's explicit vote that this risk is live. On guidance, the comparison to American Express Co is instructive: Squeri on the Q1 2026 call chose to reinvest a 10% spend-growth beat into marketing and technology rather than raise the 9-10% revenue and $17.30-$17.90 EPS range, and the stock still trades at a 22.0x P/E in the table. SoFi's premium multiple gives less room for that choice.

The asymmetric pay-off sits in the table already. SoFi Technologies at -35.0% YTD and -47.2% off the 52-week high is the worst-performing name in the credit-services peer set, deeper than Capital One Financial Corp at -14.4% and an order of magnitude below Visa Inc. at +3.0% and American Express Co at -4.7%. The 14.7% short interest is the highest in the table aside from OppFi Inc. at 3.2%, and both names carry credit-cycle exposure. If even one of the three pieces above breaks cleanly, if loan hold stops scaling, if credit provisions surprise to the upside, if the second-half cadence undershoots: the 37.8x multiple compresses against a revenue base that is no longer compounding at 23.1%. The setup is binary because the multiple is binary: a balance-sheet platform priced like a growth tech name has to deliver growth-tech execution, and the margin for missing on any single quarter is the size of the drawdown already absorbed.

Bull case wins on trajectory, bear case wins on margin of error

The earlier sections established the facts this verdict has to weigh. SoFi Technologies, Inc. closed at $17.01, -35.0% year-to-date and off 47.2% from the $32.21 52-week high, with 14.7% short interest and a 37.8x trailing P/E on $608.5M of TTM revenue](/stocks/SOFI/documents/02b9dbe0-f841-4f80-8eff-f472b7f48449). Guidance is for at least 30% member growth in 2026, adjusted net revenue of approximately $4.655B, adjusted EBITDA margin of approximately 34%, and adjusted EPS of approximately $0.60 . SoFi Bank CET1 sits at 15.3% versus a 7.0% required minimum, and tangible book value per share is up 57% year-over-year .

On the trajectory the evidence describes, the bull case wins. Demand records landed on every lending product in Q1 2026: personal loans, student loans (+2.2x year-over-year), home loans (+2.4x), brokerage fees, interchange revenue up 54% on roughly $25B of annualized card spend . Members grew 35% year-over-year to 14.7 million and products 39% to nearly 22.2 million, both already running above the guided 30% member-growth pace for the full year. Cost of funds fell faster than asset yields, and the resulting 22 basis-point NIM expansion came from a structural funding remix rather than a loan-yield story. Capital headroom is multiples above required minima and is being put to work at the balance sheet versus the loan platform. No peer in the credit-services set is posting organic revenue growth above the 23.1% in the performance table, and none is pairing it with the funding-cost improvement SoFi just printed.

On the margin of error the bear case wins. The 37.8x trailing P/E in the valuation table is more than double American Express Co's 22.0x and substantially above Mastercard Inc's 31.7x, leaving room for compression on any single quarter that misses the guided cadence. American Express Co reinvested a 10% spend-growth beat into marketing and technology rather than lift its 9-10% revenue and $17.30-$17.90 EPS range , and trades at 22.0x, a P/E that leaves room for a reinvestment year. SoFi's multiple does not. Noto's own framing on the Q1 2026 call was that this is obviously a year in which revenue is going to be disappointing because of a loss to that large customer, with a stated medium-term target of 20-25% compounding growth. If the lost-customer gap does not close in the second half, the 23.1% revenue growth in the table becomes the ceiling, and the 14.7% short base is the trade that gets paid.

The bound on the bear direction is the drawdown already absorbed. At -47.2% off the 52-week high and -35.0% year-to-date, SoFi trades at the steepest loss in the credit-services peer set, deeper than Capital One Financial Corp's -14.4% YTD and -19.8% off its high. A re-rating toward Mastercard Inc's 31.7x on the same trailing earnings would require roughly a 16% decline from the current multiple, in line with the credit-services cohort's own drawdowns rather than the catastrophic end of the bear case. The structural floor sits in the capital base: tangible book per share of $7.21 implies a book-value floor well below $17.01, and CET1 at 15.3% means capital absorption capacity does not bind on credit deterioration at the size of the current loan book. The bear case has to argue that originations, NIM, and credit all move against SoFi simultaneously to compress the multiple by more than the cohort has already absorbed.

The bound on the bull direction is the multiple already paid. Reaching Visa Inc.'s forward implied growth-adjusted multiple is not in the cards without sustained execution on the guided 30% adjusted net revenue growth and approximately 30% adjusted EBITDA margin, plus the loss-to-large-customer gap closing by year-end. The 23.1% revenue growth in the table needs to convert into the roughly 30% management has guided, member additions of at least 30% need to deliver against an already-record Q1, and the 5.94% Q1 NIM needs to defend against the 63 basis-point decline in asset yields still running through the income statement. Each of those pieces can land without the multiple rerating, and the asymmetric upside is bounded by the fact that SoFi already trades at a premium to peers on a metric, revenue scale, that lags the networks by two orders of magnitude.

What to monitor next. The Q2 2026 print, not yet on the platform as of this writing, is the next load-bearing data point: management has guided approximately 30% adjusted net revenue growth and approximately 30% adjusted EBITDA margin , and any delivery above the lower bound of the cadence would test whether the cost-of-funds tailwind is durable. The loan-versus-loan-platform mix in the Q2 disclosures will determine whether the NIM expansion is repeating at 5.94% or fading into mix. Credit metrics, early-stage delinquencies and the in-quarter net charge-off rate on the personal-loan book, will be the cleanest read on whether the 63 basis-point decline in asset yields is being earned at acceptable loss content. The SoFi Plus relaunch, the new core banking platform launch on July 1, and the SoFiUSD stablecoin that became the first issued by a US national bank to launch on a banking platform on May 27 are the cross-checking indicators on whether the platform-versus-bank framing in the bull case is being delivered.

The verdict: the trajectory evidence better supports the bull case, but the bear case has already been paid for in the table. The trade from here is a multiple-compression bet against record originations and a 22 basis-point NIM expansion, with the structural floor sitting in capital headroom and the structural ceiling sitting in the multiple peers are not being asked to pay.

Data comes from SEC filings, earnings-call transcripts, and Equibles' deterministic datasets; commentary was gathered from earnings-call transcripts and filings by Equibles research agents. This article is for information only and is not investment advice.