2023-10-05 09:38:52 ET
Summary
- Shorting Sofi Technologies above $10 makes sense due to its exuberant valuation in uncertain macro times.
- SOFI revenue growth is likely to stall in 2024.
- AFS accounting treatment creates revenue headwinds.
- It also means that SOFI has smaller buffer for credit losses should a recession ensue.
- SOFI is a highly regulated utility-like bank, it is not a high growth tech company.
I believe shorting Sofi Technologies ( SOFI ) above the $10 share price makes an abundance of sense. From a valuation perspective, this represents a valuation of more than 3x tangible book value ("TBV") which is an irrationally exuberant valuation given its ROE profile and cost of capital. Furthermore, if you adjust its accounting methodology to align with peer banks, then the valuation is even more glaring at around 4x TBV.
The bull narrative is centered around a number of lines, including:
- It is not like other banks, it has sustainable 50%+ revenue growth rates;
- It is growing deposits rapidly whereas the big banks are losing deposits;
- It is a tech company and not a bank and should be valued accordingly;
- It grows members rapidly that will be monetized over time by cross-selling;
- It will shoot to the moon once it becomes GAAP profitable.
The bear thesis is as follows:
- SOFI is predominantly an unsecured consumer finance company;
- The revenue growth print in recent years is not sustainable in my opinion and is about to slow down markedly in 2024 as it becomes capital-constrained;
- The return profile of the company once at scale is in the range of mid-teens ROE;
- It should be ascribed a relatively high cost of capital given the riskiness of its loan portfolio;
- The accounting practice selected by SOFI of Available for Sale ("AFS") for personal loans, results in upfronting of profits compared to peers. This will reverse and become a headwind in 2024;
- In a recession with high unemployment (6%+), it may need to raise fresh capital.
In this article, I will explain in more detail the bear narrative and specifically why I believe that SOFI is over-valued at its current share price, and why I short it when it trades above $10.
Note that currently SoFi's shares outstanding are ~950 million. Therefore a share price at $10 translates to a market cap of $9.5 billion or approximately 3x its tangible capital of ~$3.25 billion (see chart below on capital ratios as of 30th June 2023).
The Revenue Growth Myth
Many of the bulls would highlight the spectacular growth achieved by SOFI in recent years and attribute it to exceptional execution by the management team.
But in my view, that is clearly not an accurate narrative. The revenue growth delivered by SOFI was predominantly due to a huge increase in the unsecured personal lending space facilitated by the deployment of excess capital SOFI held.
SOFI Investor Relations SOFI Investor Relations
As you can see from above, Net Interest Income ("NII") is up 108% year-on-year predominantly driven by personal lending.
So how did SOFI manage to grow personal loans in such a significant manner? Clearly part of it is the base effect (i.e. starting from relatively low balances). However, the key reason is excess capital. SoFi simply had plenty of excess capital and it decided to deploy it to originate unsecured personal loans. Now ~68% of its loans are unsecured, yielding ~13%.
It also benefited from a lack of competition. Most of its peers were either short of capital or deleveraging whereas borrower demand for unsecured loans has been strong.
The Capital Constraint
The growth rate of warehousing loans on the balance sheet is clearly not sustainable in my opinion. This can be seen by considering SOFI's capital ratios trajectory over the last 15 months:
SoFi Capital Ratios as of 31 March 2022
SoFi Capital Ratios as of 30 June 2023
As you can see from above, SOFI has been depleting its capital ratios quite rapidly. On the basis of the back-of-envelope calculation, SOFI will be capital-constrained somewhere around 1H of 2024. In other words, growth in lending has to slow down materially.
Beneficial Accounting Treatment
SOFI selected to use AFS accounting treatment compared to peers that use Held For Investment ("HFI"). The advantage for SOFI is that it gets to recognize profits earlier compared to peers. The accounting treatment can be illustrated by the below charts:
Note the above chart is a simplified illustration of the accounting treatment and is not to any particular scale or incorporates any fair market adjustments that may occur along the way. The charts intend to demonstrate, on a high-level only, the timing difference in the recognition of accounting income between the Held-For-Investment versus AFS. For the accounting geeks amongst you, you may want to read the following guide which explains the accounting treatment in full detail , but I will summarise it below as it pertains to SoFi.
So as can be seen from above, SOFI is recognizing a day 1 profit of ~4.3% (this is Gain On Sale Margin or "GOSM") whereas peers are required to provide for loan losses of ~7% (this is in line with the CECL provisioning methodology for loans held to maturity). So the accounting difference is ~11.3% on day 1, for precisely the same loan and cashflows. This is merely a timing difference and it reverses over time but it is especially beneficial to SOFI during the ramp-up stage of lending. If I adjust SOFI accounting profits and capital to equalize to peers, this would result in a ~$800 million hit to the bottom line. The way I arrived at ~$800 was by multiplying SoFi personal loans balance as of Q3 (~$13 billion) by 6% reflecting a conservative loan loss provision (as an example, LendingClub recognized lifetime loan losses between 7% and 9%). In other words, there will be revenue headwinds for SOFI in 2024 and beyond as the loan portfolio matures.
The Return Profile & Valuation
The SoFI management team has spoken in various forums about the marginal ROE on personal loans and student loan products (40% and 20% respectively). However, these marginal ROE disclosures exclude taxes and variable costs as implied by the calculation walk-through by the CFO in the above-mentioned forum:
At the end of the day, we're looking to maximize the return on every single loan that we underwrite and we're going to do that whether that means selling or holding. On student loans specifically, to run you through the math that we've talked about publicly, similar to personal loans, if you take that 6.5% weighted average coupon that I just talked about, we're experiencing 30 basis points of annualized losses and our funding costs are about 3.85% as of Q1.
Yes, our deposit rate has gone up, but remember that's just for direct deposits and the blend is still relatively low. So, if you take those three numbers, the 6.5 minus 30 basis points, minus 3.85%, you're getting a 2.4% ROA and on a levered basis or ROE perspective, that's about a 20% ROE.
Incorporating these yields a marginal ROE of approximately mid-20s and low teens for these products respectively. This also seems to align broadly with disclosures made by LendingClub ( LC ) as per the below chart for personal loans :
But of course, these are only marginal returns and also exclude all indirect costs which are substantial. Putting this all together, on a back-of-the-envelope calculation, SOFI when fully scaled and profitable, is likely looking at an ROE profile in the range of mid-teens (or 20% if I am optimistic).
However, SOFI's cost of capital should be quite high as well given its large exposure to unsecured credit as well as being hostage to the mood swings of the capital markets should it decide to sell assets.
In terms of valuation, it appears to me that 2x TBV is a fair value that also incorporates some optionality when it comes to the Tech division as well as superior growth opportunities compared to other digital, multi-product banks.
As a point of comparison, Ally Financial ( ALLY ) is a branchless consumer finance company that focuses on auto lending, as well as various other products (mortgages, credit cards, wealth management, point of sale financing, etc.) with a much bigger scale, is generating mid-teens ROE and currently trading at ~0.8x TBV.
Risks to the Short Thesis
A key risk is that the Tech division begins to fire on all cylinders and start delivering 20-30% revenue growth. If that happens, that would be a complete game-changer.
The other obvious risk is the massive support of the stock by retail investors. This can be reflected in the stock not trading in line with fundamentals.
Finally, if SOFI manages to deliver substantial capital light fee growth (e.g. from brokerage or other products) it is able to substitute revenue growth from loan products. I see this as somewhat unlikely.
Final Thoughts
I believe that many investors are wrongly extrapolating SOFI's recent growth for the next few years. This is clearly not sustainable given upcoming capital constraints and base effect in my opinion. SOFI is likely to also face accounting headwinds given its AFS accounting treatment especially so if a recession ensues.
Investors should be cognizant that banking products are quasi-commoditized and banks themselves are essentially highly regulated utilities. Also, the one-stop-shop or financial supermarket cross-selling strategy is already utilized by various banks (online and otherwise) and with mixed success. SOFI is not reinventing the wheel here. Maintaining a collection of sub-scale products is prohibitively expensive and often not always the best use of capital.
To be clear, I am not suggesting that SOFI will not be a successful and profitable bank. At the right price and time, I could go long (as I have in the past). Rather, my short thesis is premised strictly on exuberant valuation driven by unrealistic expectations by some investors.
For further details see:
SoFi Technologies: Shorting Above $10 Makes Sense