SolStock
SoFi Technologies’ (NASDAQ:SOFI) financial releases have not been traditionally in line with disclosures I generally expect from banks and other financial institutions.
I mostly attributed it to SoFi’s self-proclaimed description of a “consumer technology company that delivers financial services products”. So, like many other investors, I have (reluctantly so) grown accustomed to considering other metrics such “Adjusted EBITDA”, “Contribution Margin” and “Product Numbers” which are more aligned with tech companies’ financial reporting frameworks.
The issue, though, is that SoFi is now behaving more and more like a bank as it is keeping loans on the balance sheet for longer, but its accounting practices are not keeping up.
The Thesis: Investors Are Not Grasping The Whole Story
The key issue is the accounting treatment of loans that SoFi is currently holding on the balance sheet.
I specifically refer to consumer personal loans for which the average balance has grown from ~$4 billion in Q2-2022 to $6.8 billion in Q3-2022. Clearly, this is a significant growth of ~$2.8 billion in 1 quarter, and therefore I was very interested in identifying the extent of loan loss provisions SOFI recognized in line with its Current Expected Credit Losses (“CECL”) provisioning methodology. For readers who are unfamiliar with CECL, it is a relatively new accounting standard that requires banks (and other companies) to recognize expected lifetime losses upfront (i.e., on day 1) as the asset is originated.
To my surprise, I could not identify the CECL number in the earnings release as well as other disclosures. So I did some more digging and found the answer. It appears that SoFi is treating its assets as Held For Sale (“HFS”) as opposed to Held For Investment (“HFI”). The former does not require SoFi to apply CECL as it intends to sell these assets before maturity and thus is presumably recognized on a mark-to-market basis.
This issue (and concern) is explained in the attached article:
With the company currently in transition to deposit-based funding, Wedbush’s David Chiaverini thinks the two major concerns’ weighing on investors’ minds and impacting sentiment are: “1) the switch to held-for-investment (HFI) accounting for SOFI’s loans from its current practice of holding loans as held-for-sale (HFS) could ultimately weigh on EPS as this switch should ultimately lead to higher loan loss provisioning (LLP) in accordance with CECL (current expected credit loss), and 2) the sustainability of gain on sale income and hedging gains is being called into question given these hedges appear to be over-earning relative to origination levels.”
Chiaverini addresses both these issues. As for the former, while the analyst admits the move to HFI accounting could “weigh” on EPS, he notes that talks with the company suggest it does not plan on moving to HFI accounting “anytime soon.” There is no restriction for the timeframe when HFS loans must be sold, and so long as a company shows it plans to sell at some point in the future, it is able to “designate” loans as HFS. During its deposit funding transition, SOFi’s plan is to lengthen the timeframe in which it holds loans – from its previous practice of holding them for 3 months to 6 months. “In theory,” notes Chiaverini, “SOFI could extend this timeframe indefinitely as long as it intends to sell the loans prior to maturity.”
The above article suggests (based on discussions with management) that the company does not intend to move to HFI accounting (as LC does). A back-of-the-envelope calculation suggests (based on comparison with LC), that if it did move to HFI, SoFi would need to increase its provisions materially, say a coverage ratio of ~7% of its $6.8 billion balance or incremental ~500 million provisions. Additionally
It seems, though, that it is holding assets for longer and longer as confirmed by the Q3-2022 earnings call transcript as opposed to selling these assets (as you would expect from HFS assets):
Kevin Barker
Just a follow-up on some of those comments around the gain on sale and everything. It looks like you originated about $2.8 billion of personal loans this quarter. And the growth on balance sheet, it was a similar amount. Is there some strategy whereby your holding all these loans or there are certain loans that are remaining on the balance sheet for an extended period of time? I know you talked about it earlier this year where you would hold them for average would push out. But is there something going on with the personal loan originations that would cause you to hold them on the balance sheet, maybe have the sales occur mostly in the fourth quarter?
Chris Lapointe
Yes. So what I would say is our strategy hasn’t changed. Given the flexibility that we now have with the bank, we’re able to hold loans for a longer period of time. And that was certainly something that we did in Q3 as well as Q2.
What I would say is having the bank provides us with sufficiently additional flexibility and a new source of funding which allows us to grow that balance sheet and hold loans for a longer period of time. There are a few ways that we can grow loans on the balance sheet and generate net interest income. We can either originate, which has been the primary source of driving volume in terms of personal loans, but we can also repurchase loans.
Clearly, as SoFi holds loans for longer, credit risks are more likely to manifest, especially if the economy goes south in 2023. The alternative would be to sell these assets down, but given current market conditions and waning investors’ appetite, that may prove to be challenging. And if it does sell the loan portfolio as the HFS designation suggests, then it will not benefit from the NII tailwinds currently baked in the quarterly earnings.
Eat The Cake And Have It Too?
NII in Q3-2022 has grown from $72m to $139m year-on-year. Whereas, non-NII income in the Lending segment has grown from $136m to $162m. So the lion’s share of the growth is from NII.
So in a nutshell, SoFi is benefiting from the NII revenue but not booking the CECL provision yet holding the loans for longer.
I suspect SoFi is likely to retain the personal lending balances high going forward, as these drive the growth in revenue for the firm. If it will sell most of these in Q4, then there would be a massive loss of income in subsequent quarters.
But What Are The Risks?
These loans will begin to season and SoFi will likely start incurring loan losses. This will be made much worse if the economy goes south and unemployment increases materially, as many expect.
The counter-argument would be that SoFi’s credit risk is not high. The portfolio has an average FICO score of 747, so credit losses should not be outsized.
Well, as a point of reference, I would contrast LendingClub’s (LC) portfolio comprising an average FICO score of 730 which is not that dissimilar to SoFi’s. LC’s provision coverage ratio is greater than 7% of loans currently. SoFi is well behind the 8-ball on recognizing loan loss provisions.
Final Thoughts
I don’t really appreciate that SoFi is reporting adjusted EBIDTA numbers as a key metric as opposed to GAAP net income. For example, share-based compensation for Q3 was ~77 million whereas the depreciation charge was greater than $40 million.
The designation as HFS and not booking CECL provisions, even though it now expects to hold the loans for longer is clearly within the blackletter law and an acceptable accounting practice, but I am not convinced retail investors transparently understand the full implications when assessing SoFi’s “adjusted EBIDTA” metric. As a point of reference, If I extrapolate SoFi methodology to LC’s financials, one should add ~50m to LC’s net income in Q3 alone. Note that LC reports on a full GAAP net income basis, including share-based compensation and depreciation costs. It also designates assets as HFI and provides conservatively for lifetime expected losses. This is not the case with SoFi. So, from my perspective only, I would take their financial disclosures with a grain of salt.
Having said that, I also do acknowledge some of the good progress SoFi has made, especially in the context of deposit raising and lowering its cost of funds.
I currently hold a small position in SoFi. I do not intend to increase this in the short term. I suspect, at some point in the cycle, SoFi will experience outsized and unexpected credit losses.
Be the first to comment