SoFi is a one-stop shop for financial services. It aims to provide consumers with excellent products for all major needs and events. It further pushes to delight them to the point of never going anywhere else. That creates key cross-selling and acquisition cost benefits. It also owns its tech stack, has no branches and possesses a bank charter, which all give it the rare ability to combine the cost edges that incumbents and fintechs each enjoy. It also sells its tech stack to customers like H&R Block and Wyndham Hotels. It has a large lending business with an increasingly diverse array of funding options, as well as bank accounts, brokerage services, 3rd-party insurance options, credit cards and so much more.

My SoFi Deep Dive can be found here. This gets into intricate detail on SoFi’s unique value proposition within banking, and everything else you need to know about the firm.

It is great to see product growth moving from trailing member growth on a 2-year basis to matching it in the Y/Y growth column. That shows increasingly healthy engagement for newer members… and that trend is accelerating with slightly faster Q/Q product growth. An incremental 1.5 products added per member Q/Q is a very strong result for this business. As we’ll explore in more detail later on, this is the “better together” cross-selling machine in action.

The tech platform Q/Q decline in accounts reflects a client that diversified processing vendors last year. That was already in the 2025 guide and there are no changes to its expectations for that segment for 2025 or 2026.

Fair Value Accounting Refresher:

As we work through this section, recall that net charge-off is a lagging credit indicator, while delinquency rates are more of a leading indicator. Also note that SoFi doesn’t practice current expected credit loss (CECL) accounting like most of its counterparts. Instead, based on its desire to frequently sell loans into capital markets, it uses fair value accounting. This leans on an independent 3rd party auditor and highly conservative macro assumptions to set fair value markings on loan pools every quarter. To avoid pent-up unrealized losses (big part of the regional banking crisis), changes in fair value flow through the income statement every single quarter. It also hedges away all of these fair value changes to eliminate the conflict of interest coming from unfairly propping up markings. For these reasons, while I am in the minority, I think fair value is superior to CECL.

Capital Market Activity:

SoFi had a banner quarter for capital market transactions, which include loan pool sales and the lending platform business (LPB). Specifically, it sold or transferred $3.1B in personal loan volume, representing a quarterly record and triple the volume of any quarter over the past year. Even if we excluded LPB, its whole loan sales matched its highest rate in a year. And to make things more encouraging, personal whole loan sales were again executed at gain on sale margin in excess of its fair value markings. Specifically, the price tag was 106.2% vs. its fair value marking of 105.5%

I’ve said it before and I’ll say it again. Critics of SoFi fair value accounting ignore its ultra-prime demographic, its write-down practices and its cross-cycle underwriting performance. They turn their nose up at all of those tangible items to conclude fair value markings are unfair. Well? Those people don’t have anywhere near the detail or quality of data that capital market buyers have access to – including within LPB. And? These capital market buyers are telling us they’re willing to pay more than the perceived fair value. Every. Single. Quarter.

Notably, SoFi also sold nearly $700M in securitized loans previously originated for partners through LPB. This deal structure hasn’t happened in 4 years, and just shows us that capital markets remain wide open for this company.

“This channel provides our partners with meaningful liquidity to support their ongoing investment in LPB. The transaction priced at industry-leading cost-of-funds levels… with an all-in yield of 5.10%.”

CFO Chris Lapointe

It sold $322M in home loans at a gain on sale margin of 102.1%. 33% of this volume was from home equity loans, which is a highly popular product for consumers locked into very low-rate mortgages. It sees volumes here accelerating throughout 2025.

“I thought some of our [capital market] deals would get delayed when the tariff and volatility stuff happened… There was no slowdown…  People are moving just as fast and it was a pleasant surprise.”

CEO Anthony Noto

Credit Health:

We again got some very helpful and encouraging cohort-level (or vintage-level) analysis for SoFi’s credit health. For loan pools from Q1 2022 - Q2 2024, its 4.09% cumulative loss rate is far below the 5.53% it reached in 2017 when it last approached its life of loan loss rate tolerance of 7%-8%. This gap widened by 16 bps Q/Q after widening by 15 bps Q/Q last quarter, as newer vintages continue to perform better and better. For pools ranging from Q1 2022 - Q4 2024, 59% of principal is paid down with a 6.7% cumulative loss rate. Losses would have to reach 10% to eclipse its loss rate target, which has never happened. This makes it crystal clear why capital market buyers continue to welcome SoFi with open arms.

The firm sold $90 million in late-stage delinquent loans during the quarter for positive value realization vs. waiting for them to be charged-off. The bottom row in the chart below accounts for that, and still shows more improvement without these (entirely kosher) items.

More Balance Sheet & Capital Ratio Data:

As you can see below, SoFi’s capital ratio cushion has begun to shrink as expected, with its well-communicated acceleration in lending activity.

More Student Loan Metrics:

For the full year, SoFi raised revenue guidance by 1%, which beat estimates by 2.6%. It raised EBITDA guidance by 3.5%, which beat by 3.1% and also raised $0.26 EPS guidance by $0.015, which beat by $0.025. The $35M raise to revenue guidance is slightly smaller than the Q1 beat. The $30M raise to EBITDA guidance is in line with the Q1 beat. I think this prudence is absolutely the right decision in this backdrop, and it’s a very easy pill to swallow when it still coincides with an annual raise. Tax rate assumption remained at 26%, so profit raises were fortunately not helped by that. It also now expects to add $592.5 million in tangible book value vs. $562.5 million based on its previous guidance. Finally, it reiterated at least 28% Y/Y member growth.

Q2 was a bit ahead on revenue and EBITDA and in line on GAAP EPS.

SoFi trades for 50x forward EPS. EPS is expected to compound at a 2-year clip of 75%. I expect profit estimates to very modestly rise following this report.

Lending Platform Business Refresher (Review for Consistent Readers):

As a reminder, SoFi has a lending segment, SoFi Lantern and a Loan Platform Business (LPB). Lending revenue comes from originating loans for its own balance sheet and either holding them or selling them to capital market buyers. That’s lending revenue. When borrowers are rejected, SoFi can send applicants to its loan marketplace, called Lantern, to match them with other creditors in exchange for a referral fee. That’s financial service revenue. More recently, as of Q3 2024, it now signs agreements with partners like Fortress Capital to originate loans on their behalf – right from SoFi’s app or site. This is for both borrowers within its credit bands, if it doesn’t have the volume capacity to originate, and outside of its credit bands too. Here, servicing revenue is part of its lending bucket with everything else (origination & other up-front fees) part of financial services revenue.

Asset-Light Demand — LPB:

Over the last few years, SoFi has been optimizing its balance sheet with the banking charter in hand. Simply put, this just means it’s matching the lowest cost assets it can with the highest interest products it can provide. This has boosted revenue growth considerably, as it has been able to generate considerably more net interest income on its book of business. It has also allowed SoFi to pocket nearly 200 bps in profit spread from originations or over $500M in annual cost savings.

That process has essentially wrapped up, which makes revenue growth outside of balance sheet optimization vital for sustainable compounding going forward. Fee revenue is what to focus on. It includes financial services products such as lending origination fees, referral fees for things like insurance, interchange fees, brokerage fees and also LPB. 

Not only does fee revenue come with more durable growth, but within LPB specifically, it allows SoFi to untether borrower servicing from taking more balance sheet risk. It lets the company tap into eager capital market suppliers to take that balance sheet risk themselves, while it collects high-margin, low-risk revenue. Asset-light, fee-based businesses are routinely rewarded with a higher valuation multiple than a traditional lending shop. Investors understandably love higher asset turnover, higher asset productivity, and less risk.

Fee-based revenue rose 67% Y/Y to $315M, as it crossed 40% of company-wide revenue vs. 27% a little over a year ago. It’s now at an annualized rate of $1.3 billion, which is about 50% of total 2024 revenue. Quite the rapid and positive shift… and it will keep shifting, per the team. LPB is basically a soon-to-be larger, faster-growing Upstart (UPST) tucked into a full suite of SoFi services. But? Fortunately with an ultra-prime niche rather than a vicarious subprime niche. 

For LPB specifically, the product did $1.6 billion in total originations and $96 million in total revenue. They expect this to accelerate throughout 2025 (despite bond market volatility). Excluding the small piece of revenue attributable to its lending segment, this equates to a 5.8% take rate for LPB this quarter. Not only does this mark 45% sequential growth for LPB, but another quarter of a take rate nearing 6%. They’re at a projected annualized volume of $6B and revenue of $380M, which represents a 6.3% take rate “due to excess demand.” In my previous projections for how much revenue LPB deals could generate, I assumed a conservative 4% take rate. That appears to be way too low (a very good thing). It has racked up $8B in new commitments for LPB, between aforementioned deals with Blue Owl, and others, to keep this growth engine humming.

Most of its LPB volume originated for 3rd parties continues to be within its ultra-prime credit demographic. Lantern is more for subprime borrowers referred to others. SoFi rejects about 5 borrowers per every consumer it approves, and Lantern helps it route those rejected applicants to other options (while getting them in the SoFi ecosystem). It sounds like LPB will try to expand to wider credit bands in the future, but I’d candidly love for them to focus on ultra-prime. It’s far less cyclical and clearly a gigantic opportunity.

More on Financial Services:

Interchange revenue soared 90% Y/Y to reach $14 billion in annualized spend ($16 billion including credit interchange revenue), while financial services provision for credit losses fell 21% Y/Y. Rapid processing growth paired with lower credit loss projections is a great combination. SoFi Relay (tracks assets from SoFi and outside of it on 1 dashboard) enjoyed 41% Y/Y product growth, while 33% of SoFI Relay customers opened a 2nd SoFi product within a month. Relay doesn’t directly make money, so the cross-selling note is highly important.

Net Interest Income & Pricing Power:

It is good to see SoF’s cost of funds falling faster than average asset yields. This shows SoFi can easily cut deposit rates and access other low-cost funding pools to maintain (or grow) net interest margin (NIM) in a falling rate environment. This happening while deposit growth remains as strong as it is shows us great elasticity of demand for SoFi’s business. It continues to expect its APY to remain among the top in the market, as it doesn’t rely on mandated APY levels from sponsor banks like the non-chartered fintechs offering competing products. If it needed to, it could offer the highest APY on the market today, but deposit levels are excellent at this rate, so there’s no motivation to do so.

“We continue to expect a healthy net interest margin above 5% for the foreseeable future.”

CFO Chris Lapointe

More on Lending:

Loan volume originations rose 69% Y/Y or 31% Y/Y excluding LPB, which predominantly presides in the financial services bucket. Student loan volume rose 59% Y/Y and home loan volume rose 54% Y/Y, as its integration. Wyndham Capital is providing the expected stable backend foundation to support faster originations and acceptable service scores. It will soon debut a “new revolving personal loan product for prime credit card customers,” which is meant to undercut existing products. It can save borrowers up to 40% on payments, while still generating a 30% return on equity for SoFi.

Tech Platform:

This remains the most challenging part of the business. It is still getting through a painful change in go-to-market to prioritize larger, higher-quality customers. That change coincided with macro souring, a regional banking crisis and a massive Technisys integration project. It’s fair to call all of these excuses, but I think they’re fair excuses… especially when they all came at the same time.

SoFi Plus Subscription:

SoFi Plus is driving great cross-sell activity. For existing SoFi members adding Sofi Plus (like yours truly), 30% are taking out another product within a month. For SoFi Plus members who are brand new to the ecosystem, 75% are taking out a 2nd product within 30 days and 40% are taking out a 3rd. This was a key part of 32% of all new products being opened by existing members and net new products eclipsing net new members by 50%. Working as intended.

Brand-Building Investments:

Investments and promotional events across broadcasting, social media and more for the “Tomorrow’s Golf League” (TGL) sponsorship went very well. This racked up 20 million total viewers, with a SoFi Invest promotion that led to “record engagement” for that product.

SoFi Money:

“We are now the only company that offers digital person-to-person payments via phone number or email address and the ability to send money via Zelle, ACH or self-serve wires. Given the performance of SoFi Money, we expect it to become our second one billion dollar revenue business.” – CEO Anthony Noto

SoFi Invest:

21% Y/Y product growth for SoFi Invest is its best result in a while. It’s the byproduct of an improved, cleaner user interface, as well as partnerships with Templum to expand asset class availability. This is how SoFi users will gain access to Anthropic. Beyond that, Noto hopes to add level 1 options, crypto and blockchain products by the end of 2025. None of those products are currently part of guidance and could provide modest upside.

“Given the evolving regulatory landscape, we see an opportunity to re-enter the crypto and blockchain business more comprehensively…  In addition to enabling members to invest in crypto coins, we will enter other areas over the next 6 to 24 months, but potentially much sooner via acquisition or if the changing regulatory landscape allows.”

CEO Anthony Noto

It expects SoFi Invest product growth to accelerate throughout the year, as it continues to prioritize this more than it has before. All of this work is clearly going well and record product engagement this quarter merely offers more evidence of that being the case. This was the most focus placed on SoFi Invest product innovation since it went public. Considering this market is gigantic, SoFi has done very little to advertise outside of its existing user base and SoFi Invest is 50% under monetized, this priority is well placed.

“We're accelerating our product investment to improve the Invest UX, drive more personalization and create better discovery of our Invest product offerings within our app… The more we service our great investment selection, the more engagement we see.”

CEO Anthony Noto

SoFi Protect (Insurance) a Future Focus:

“We're incubating and beginning to scale our SoFi Protect offering across home, auto and life insurance, lending and checking products for small and medium businesses.”

CEO Anthony Noto

Velocity of Innovation:

Noto echoed the comments he made on social media a few weeks ago following criticism from some investors:

“We are stepping on the gas to launch new products faster and iterate to improve our existing products at an even more rapid pace.”

CEO Anthony Noto

Finally, SoFi teased its cash coach AI tool (nudge best use of funds) that is coming soon.

Excellent quarter. I’m sure there will be people who pick on technology accounts or some other obscure data point, but they’re missing the forest for the trees.

SoFi continues to show everyone a uniquely consistent ability to compound across cycles. That’s not normal for fintech, and is a byproduct of the fantastic market share gains it continues to enjoy. Capital markets remain wide open for this company despite bond market and credit spread drama… and that’s because of the pristine reputation SoFi has built with these savvy buyers.

More rapid compounding… more credit health improvement… more surgical balance sheet management… more operating leverage… more beats and raises… more acceleration for member adds… more of the same. Go SoFi go.

I am not adding to my already large position, as it is still relatively more vulnerable to macro than most of the software-based names that I own. At the same time, I have no interest in trimming and am very happy to have what is admittedly a high beta name near the top of my portfolio. I like the amount that I own today. They’re killing it.