Ok, What the Heck is Sofi Doing in 2024? (SoFi for Dummies)

Written by Timothy Sweeney

This will be the first article published on DDI that isn’t written by me. This was written by Timothy Sweeney, who is probably smarter than me and definitely has way more experience than I do. He begins by providing his background, so I won’t repeat that here. I suggest that everyone follow him on X @Tim_Sweeney_TAR. He had an excellent post there that gained a whole lot of traction because it’s so comprehensive and well written. I asked his permission to post it here since the lifespan of a post on X is pretty short and hopefully it will make it easier for others to find later if I post it here as well. He added a few details and gave me permission to publish it. Without further ado, here is his analysis:

BACKGROUND

There are a lot of misconceptions and misunderstandings, intentional and otherwise, on X and YouTube regarding Sofi operations and results, as well as speed of new product implementation and stock price guidance.  Most of these misunderstandings come from individuals without any significant business or banking experience, and even more with barely any understanding of GAAP accounting or accounting rules.

For your information, I was a partner in a large national law firm, a partner in a large national accounting firm, a managing director of a large wealth management firm and a C-Suite executive in a fortune 500 company and a key leader in the restructuring of that company.  I have led or participated in a large number of restructurings and own several small companies.  I consult on corporate restructuring, Board of Director issues and initiatives and am an expert in valuations and forensic accounting. For example, I reconstructed 20 years of financials of a number of foreign banks and translated those resulting financials into US GAAP. I also taught accounting courses at University of Michigan for 3 years.

That being said, this post is not financial advice and you should always vet facts and opinions you read on X or hear on YouTube, and conduct your own due diligence. As I stated to someone, I am an attorney, but I’m not your attorney. I am an expert, but I’m not your expert. I’m not going to tell you to buy the stock or sell the stock.

For the record, I do own a position in Sofi, and in this post, I am just giving you some of my insights as to what I think is going on. I’m not going into the numbers – there are a number of people on this platform that do a great job of tracking and analyzing the numbers.  I just want to give you an idea of what’s likely going on in the Sofi recent developments. I will simplify it to an extent for ease of understanding .

TRANSITIONING TO A LARGE BANK

Sofi stated in their recent filings and earnings reports that is facing a transitional year where financial services and tech should become 50% of the business. So, let’s think about how you would transition from a student lending company to a major bank. Sorry to disappoint some people, but that’s simply not going to happen in 3 or 4 years or even 10 years. It’s a business journey that requires a lot of work by a lot of people that’s not shown on any chart.

Anthony Noto became Sofi’s CEO in February 2018 (when Sofi was a student lending company), and shortly after expanded Sofi operations to include other loans, checking and savings and investing with the goal of becoming the “AWS of fintech.”

ROADBLOCKS

Then Covid came and in March 2020, student loan payments were paused (which continues to this day). But they overcame that by leaning in on personal loans. Then Sofi went public in June 2021 by merging with a SPAC (and all the built-in capital problems and legacy accounting issues associated with them).

In January and February 2022, Sofi was able to garner a national banking license. Then in March 2022, the Fed stated a historical ramp up in interest rates, which typically leads to poor bank performance (and resulted in bank failures and banking stress generally). In less than 2 years after obtaining its banking license, Sofi has now had 2 profitable quarters.  During those 2 years, Sofi grew its revenue and more importantly, its members substantially (more on that later).

I always thought of Sofi as a long-term investment and as a disruptor of banking. Aside from the AWS of fintech arguments, which I knew from my experience with software implementations, could take a long time, I simply liked the fact that I didn’t have to pay account fees, check fees, overdraft fees etc. and can pay my credit card in the app.  That alone differentiates it from all my bank accounts where the bank would hold my deposits and then charge me an account fee because I did not maintain the required balance. I will never go back.

GROWING AND RESTRUCTURING SOFI

Sofi began its transition by expanding offerings and growing members. They did that and continue to do that very successfully. They needed to add new products and expertise. They have and continue to do that successfully.

However, they had a lot of capital structure issues as I reviewed their financials last November, some of which were related to the SPAC origination and others because of the huge growth in personal loans which used up a lot of their capital ratios.

There were numerous bear arguments on X saying Sofi faced a dire capital ratio problem. But all they needed to do actually was restructure their capital. I posted this on X last October/ November saying they could easily boost their capital ratios significantly by simply buying back some of their 2026 convertible debt at a discount. For every S100 million they bought back at the then 30% market discount, they would have a $30 million gain. If they bought back $1 billion of the debt, they would have an 8x or 2.4 billion increase in available capital lending capacity. The retirement of that debt at a 30% discount also lowered future potential dilution and fully diluted shares. Several so called “investment experts” disagreed.  THEY WERE WRONG.

Then I also advised Sofi needed to get rid of the preferred shares paying 12% cumulative dividend, which for accounting purposes was treated as debt and interest expense.  With the jump in interest rates on that preferred and being treated as interest, that multimillion dollar hit to eps every quarter for years needed to be eliminated.

In their latest restructuring, they converted more of the 2026 convertible debt at a discount. It was at a higher conversion ratio because the discount had fallen from 30% to 15%... and likely because it was necessary to keep their credit options open.  Because those noteholders lost so much money on that investment and Sofi was also issuing similar new notes, part of that extra dilution was likely a necessary negotiation to ensure a good market for the new convertible bonds. Although there was share dilution, there was minimal or no book value dilution.

The new 2029 convertible bond issuance was used to redeem the preferred with greater than 15% preferred dividend rate and effectively replace it with new convertible 1% debt which reduce expenses and cash flow by over $59 million a year, which not only boosts cash and future book value but eps for years.

The 2029 notes also have a cost – they also have a conversion feature and possible dilution. Part of the dilution is protected by a capped call. The issuance of 2029 notes however likely resulted in shorting of Sofi stock related to delta hedging. An example of that would be buying the convertible debt, giving you an option of acquiring the shares, then shorting the stock. If the stock goes down, your investment in the convertible is protected and you can also have gains from the short, and if the stock goes up, you can deliver the converted shares. Either way, you receive the proceeds of the short sale and can invest them and get additional returns. Sofi can also use cash to redeem the 2029 convertible debt under certain circumstances. If the stock trades sideways for a year or even two, Sofi could have an opportunity to buy some of it back at a discount also.

Will this result in dilution? Very likely, but not all dilution is bad. You obviously have dilution in shares, but whether any dilution is bad depends on how you use the proceeds to grow the company and what returns you achieve.

So now Sofi’s remaining capital issues are redeeming the $500 million balance of the 2026 notes, likely for cash at maturity in 2+ years.  They have plenty of cash, and in those 2 years, Sofi will grow book value by way more than that. The 2029 notes are due in 5 years. As I stated there are many options, depending on what happens between now and then, but most likely there will be some dilution.

WHY THE INTENTIONAL DECREASE IN LENDING?

Sofi stated that lending will be lower this year, 92% to 95% of last year, and they were tightening their lending guidelines and credit standards and even eliminating loans at lower credit tiers. Everyone wants to know why. Well, it’s not because they can't ramp up loans. Noto stated that they had that option depending on market conditions. (I would like to point out that all this was determined prior to Powell stating that he did not see the Fed raising rates this year so look for Sofi to beat their guidance again for 2nd quarter.)

The market for loans is fairly good, as verified by Sofi competitors and big banks. The reason they expect reduced lending is because they are tightening credit standards. So, you have to ask yourself why? Diehard bears would suggest that their defaults will sky rocket. But even in 2008 and thereafter, personal loans with high fico scores had very low default rates.

THE ANSWER IS IN FAIR VALUE ACCOUNTING

Keeping it simple, under fair value accounting, you can originate loans for sale or you can originate loans for investment.  With Sofi’s history of originating and selling those loans, they adopted held for sale accounting. Under that method, you originate and package loans and value them for sale based on the price you can sell them. For Sofi personal loans, as an example, because the loans are of higher credit quality, you could sell them for a premium, say at 104% to 105%. That means that if the market for 14% personal loans is par or 100, Sofi can sell them for 4% to 5% more because of their lower likely default rates than other personal loans being sold, and that calculation is also based on certain benchmark rates on risk free assets and present value rates.

Sofi would hold the loans for sale for a short period after valuing them and sell them for about that same value and recognize the premium as income immediately. That is basically fair value accounting on a held for sale basis, you determine what the value of the loans are and recognize the fair value as premiums even before you sell them. That was and is Sofi’s accounting method.

The other accounting method is CECL (current expected credit losses). The CECL impairment model requires the immediate recognition of estimated expected credit losses over the life of the financial instrument by setting up an an expense and an allowance account. You then adjust that account periodically as you determine defaults, based also on certain benchmark rates on risk free assets and present value rates.

Under HFS accounting you recognizes all the income immediately and adjust higher and lower depending on defaults and market conditions. In HFI accounting, you estimate the defaults and recognize that expense immediately, and then adjust that periodically on basically the same data.

I didn't originally want to explain this because its confusing. CECL accounting and Fair Value accounting are two separate accounting principles. It is difficult but not impossible to change methods. But what people don't understand is that Fair Value accounting under hold for investment (HFI) requires the use of a cecl impairment model. There some other rules, but my explanations herein would not change materially.

Over the term of the loans, the interest income and default expenses should be almost the same. They are both acceptable accounting methods. So, what is the issue? Because Sofi loan standards are more restrictive, its loans are valued higher than other loans and have high fair values. Sofi would recognize the fair values as income, and those loans would then be held on the books with those fair values.  Every period, Sofi then has to value those loans and adjust those fair values based on default rates, benchmark interest rates, discount rates etc.

When interest rates rose, Sofi loans had high fair values, which they would typically sell and gain that premium. They would use those loan sales as part of the evidence for their fair value calculation for loans held for sale in inventory. But Sofi determined with the high rates and good credit ratings on those loans, it could actually make more by holding those loans than selling them, and they did.

As rates continued to rise, they built up more and more loan inventory with low default rates that increased profitability. Thus, they built up substantial fair value premium adjustments on those loans. Those loan premiums (i.e. the fair values) must be amortized over the terms of the loans reducing income.  Between January 1, 2022 and December 31, 2023, Sofi personal loans on their books went from roughly $3 billion with cumulative fair adjustments of $95 million to $15 billion and with $718 million with cumulative fair adjustments.

So that’s the issue. Some analysts don’t understand why fed rates can go up and Sofi can maintain their fair values, arguing that Sofi would have to continue growing the personal loans to save it from a big write off of lowering fair values and Sofi would run out of capital ratio room to do it (which was fixed by their capital restructurings).  They also emphasized that Sofi used a similar discount rate to the risk-free fed rate and that didn’t make sense.

One alternative is to value the loans based on the present value of cash flows using the loan rate and discounting that income flow based on a standard benchmark that takes into account average default rates, which would not be anywhere close to the fed risk-free rates. But Sofi’s loans were way better than the average default rates and credit quality of those benchmarks, so it chose to determine its fair value calculation by, for example, taking 15% loans and subtracting the 4.4 % cost of capital (the blended rate it pays on deposits) and its 4% default rates, netting 6.6 % a year, then discounting those cash flows using a much closer rate to the fed’s risk-free rate. Why? Because they lowered the amount of cash flow by the defaults (and likely prepayment risk), thus risk adjusting the returns so they needed to risk adjust the discount rate close to a risk-free return.  So that argument is easily dismantled.

The next argument was that all those old loans had to have lower fair values than the new loans, so the average fair values of the entire loan book should be lower, requiring large write offs.  But as you originate more loans, the higher fair values of the new loans in greater amounts offset the lower fair values , if any, of the older loans in smaller amounts. Furthermore, a lot of those fair values on the older loans gets reduced through amortization each period and the 6.6% annual return of a much larger loan base can easily covers the say 1-1.5% amortized premiums of the older lower loan amounts in a year.  Also, on an ongoing basis, the older loans are paid and fall off of the loan book, so as rates rise, it doesn’t vary that much.

Is there room for someone using fair value accounting to take advantage of the rules and keep fair values higher? Yes. The calculation of values under fair value accounting is more complicated and has more inputs than I have described, but I am trying to keep it simple. Those fair values are reviewed and calculated by a third party valuation firm and approved by Sofi outside attorneys and accountants.

WHAT IS THE REAL FAIR VALUE RISK?

The fair value risk for Sofi related to running out of capital ratio space is gone now. They have plenty of capital ratio space and they are growing more with increased EBITDA each quarter. There were always capital options for them. The real risk is the $718 million of fair value adjustments in inventory at the end of the 2023 year. Sofi can’t let that inventory grow further.  It is a long-time burden against earnings and thus capital.

What could trigger a large write-off of this balance and a hit to the profitability trend? Well, it has always been the same things. First, default rates increasing substantially.  As we discussed, higher average default rates lower fair values.  However, Sofi default rates fell this last quarter because they presold some loans that would otherwise have defaulted, and they obtained double the return. Bears may not like it, but they can do that – it’s perfectly ok. They can also buy credit default spreads to protect against default rates (like they did for their student loans). This is done to MANAGE the fair values and is fine to do. Banks manage defaults and have entire departments to avoid defaults, modify loans and workout loans.

The other thing that could trigger fair value write offs is rapidly declining rates. Surprised? Well, here’s that issue. When rates decline, debtors will refinance those loans. When that happens, all of the fair value of that loan is written off because its then paid off.  Sofi markets the member’s ability to refinance these loans at no cost. So, if rates fall every month, a large portion of Sofi loans could refinance and cause a very large write off of those fair values in one quarter, reducing its capital necessary to grow members by taking advantage of other non-clients that want to refinance with Sofi.  Now that there is only the prospect of one rate reduction, later in the year, Sofi has an opportunity. (also, to the extent rates fall, if Sofi keeps their cost of capital (savings rates of 4.6%), their credit spreads will fall. However, they could make some of that up as the loans on the book increase in value as rates decline).

So, as a good manager that foresees this, what would you do? Well, you would start trying to reduce (let run off) those fair value premiums in inventory, which they did in the first quarter from $718 million to $608 million a 15% quarter to quarter reduction. Eps would have been roughly 10 cents a share higher on a fully diluted basis without that reduction.

You can expect that will continue for the next 3 quarters. Why? My guess is Sofi wants to run off all those fair value so it has a lot of capital ratio room to refinance its members and substantially grow members in the declining rate environment later in the year. But more importantly, I predict that after that is done, Sofi can then switch from held for investment fair value accounting with all loans being held for sale to a plan to start originating loans for investment using held for investment applying cecl impairment methods and originate other loans for sale using fair value accounting to balance their income in volatile periods, like diversifying, without the ongoing fair value overhang.

However, by running off those fair values, you need substantial origination of new loans to offset that amortization expense and maintain profitability. But if you keep the same credit box and underwriting, originating more loans that won’t sell for higher fair values, that builds up additional loans and fair value adjustments in inventory. If defaults increase, your fair value adjustments in inventory are at risk of write off.

To offset that, you would restrict credit so you can originate loans with greater credit quality that could be sold at higher premiums now (or even later if rates fall). This not only increases the average fair values on sales, but it also lowers the average defaults in inventory if you hold some of those higher quality loans. Sofi did that in the first quarter by selling $1.9 billion of personal loans at higher fair values, offsetting the $110 million of amortization.  The key is to package higher quality loans and sell most of them, and manage your defaults while you do it.

By restricting credit, Sofi naturally would likely model a decrease in lending, but it has plenty of room on the margin to increase originations without a significant risk of this plan, especially now that Powell rules out a rate hike. Furthermore, Sofi guided the roughly the same numbers for the first quarter and beat those, so it's likely they beat their guide by a similar margin in the 2nd quarter.  Their guide was just conservative, as Noto stated. Remember, he said this BEFORE Powell ruled out rate hikes. They can always take advantage of originating more loans and beat the guide substantially.

So, towards the end of the year, Sofi should have a solid balance sheet, a solid capital structure. large capital ratios and a cash position to reduce reliance on more expensive warehouse facilities, more substantially ramp up members and capture a solid share of the upcoming refinancing business in a declining rate environment, and also start originating loans at higher than market rates and selling loans they originate for higher premiums after a shorter holding period because interest rates will likely be falling.  All this without any fair value overhang. All without any fair value bear thesis. Thus, they will be transitioning to a capital and loan structure similar to a big bank.

A FEW THOUGHTS ABOUT TECH AND PRODUCTS

Regarding tech, as I stated before, those projects take about 3 years. I believe Citi and Wells Fargo and I think B of A all need new tech stacks as well as a lot of larger regional banks. They all move slowly. It’s not an easy decision, but not a lot of providers have an offering of a fully integrated stack.  Sofi should win their share of those, but certainly not all. If they get 20% gains without those wins this year, that's totally acceptable. They've only had Galileo and Technysis a short period of time.

Regarding rolling out new products... Zelle is here, but what about better user interface, Level 1 options, business banking, better credit card offering,  better etfs, asset management, tax software,  better bill pay and accounting offering, insurance, growing members 1 million per quarter etc. etc. It's likely that one or 2 of those will be this year.  If you review Sofi employment job opportunities, you'll see what they are working on. A company will typically first adopt new products that do not require a lot of integration, do not have substantial costs, that also satisfy customer needs at the most "bang for the buck." That's why a month ago I replied that you should look for Zelle soon.

You can’t do all of these new products at once or even in a year or two. They need to hire people and develop additional infrastructure for each of these. So, feel free to be bearish. It's possible the stock will stay between 7 and 9 most of this year, which is what I posted on X, but maybe higher at the end of the year.

Think of this, if they have 8 cents profitability for this whole year while running off 10 cents a quarter of fair value expense amortization, that's about 48 cents (without significant further fair value write offs) and flat 2025 guidance. Now that $.80 guidance for 2026 seems to be a bit low, doesn’t it?

Add to that, what happens if Sofi gets to 30 million members and develops small fee based premium credit card that they provide to half those members and earn net $40 profit. That’s $600 million more in net income. What if they provide asset management and custodial fees, that could be even more. My conclusion is…IT’S ALL ABOUT THE MEMBERS.

I'm investing for 5+ years for what I see is their future. I have a lot of legal, accounting, and restructuring experience. And as long as they are proceeding based on what I think is the proper strategy and making progress, I'll invest more, especially at these prices. We’ll all have to wait and see.

Sofi is making great progress.