Upstart research by YieldFanatic

Hi all,

One of the Twitter accounts I follow, YieldFanatic, which provides thoughtful content, has released an analysis on upstart.

For now, his service is entirely free, so you can find the presentation here:

https://c4681236-adf2-4da7-ba41-52b9c0b56b3d.filesusr.com/ug…

(The links works for me, hopefully it does for you)

TL;DR version: he is not bullish on the stock at all. I don’t agree with many of his points (overhyped AI, many risks) but interesting !

9 Likes

I read through most of the Presentation but honestly couldn’t finish it. I do agree it is a good aggregation of information regarding Upstart but their conclusions based on that information are really poor. Case in point: they continually compared Upstart to The Lending Club. Upstart is a software provider for the assessment of credit risk while The Lending Club is a loan originator…huge difference. From the Lending Club’s 10-K:

With respect to investors, we primarily compete with other investment vehicles and asset classes, such as equities, bonds and short-term fixed income securities.

Also I would add that one interview with Paul Gu and you would see that Upstart’s AI claims are not just marketing hype. This certainly doesn’t make me want to look further into the Yield Fanatic service.

16 Likes

I want to add one more thing that this person harps on. They mention multiple times this “Repurchase Obligation” and how great of a risk it is. They make it seem as if Upstart must repurchase any loan at any time if the institution or originating bank is not happy. From Upstart’s 10-K I believe this is inaccurate and makes me think Yield Fanatic has a short position. All quotes below are from the 10-K.

"Repurchase and Indemnification Contingency: Under the terms of the loan purchase and loan servicing agreements between the Company and institutional investors, as well as in agreements with investors in securitizations where the Company is not the sponsor, the Company may, in certain circumstances, become obligated to repurchase loans from such investors. Generally, these circumstances include the occurrence of verifiable identity theft, the failure of sold loans to meet the terms of certain loan-level representations and warranties that speak as of the time of origination or sale, the failure to comply with other contractual terms with the investors, or a violation of the applicable federal, state, or local lending laws."

"In our loan funding programs, including asset-backed securitizations and whole loan sales, we make numerous representations and warranties concerning the characteristics of the Upstart-powered loans sold and transferred in connection with such transactions, including representations and warranties that the loans meet the eligibility requirements of those facilities and of investors in our loan funding programs. If those representations and warranties were not accurate when made, we may be required to repurchase the underlying loans. Failure to repurchase so-called ineligible loans when required could constitute an event of default or termination event under the agreements governing our various loan funding programs. Through December 31, 2020, the number of repurchased Upstart-powered loans as a result of inaccurate representations and warranties represents less than 0.34% of all Upstart-powered loans."

The conclusion from above: if Upstart provides information that is inaccurate, they must buy back the fraudulent loan that was issued.

I tend to get aggravated when in this age of the internet, anyone can say whatever they please with no review or accountability. Twisting the facts for your personal gain is wrong.

If I am wrong in my conclusions please correct me!

-Junomean2
Long UPST

24 Likes

I agree with junomean2.

The writeup is good at describing the mechanics of how all the fees inter-relate: the customer pays the 0-8% origination fee to Upstart and Upstart passes that fee to the bank, then the bank pays Upstart the 3-4% referral fee and the 2% platform fee.

The writeup is also good at mentioning the mechanics of how the loans get securitized: the banks sell most of the loans back to Upstart and then Upstart securitizes them. The writeup is good at pointing out how Upstart depends on the securitization market. I feel better about this risk after seeing Upstart navigate the COVID crisis.

The writeup is good at pointing out how the bank concentration with CRB is overplayed.

The writeup is good about pointing out the tail risk of approximately $6 billion in rep. and warranty provisions Upstarts indemnifies, and also good at pointing out that this is not a large risk (he gives it a “two red flags” out of his five red flag rating system). We last saw rep. and war. in the news in the 2008 financial crisis, where the stated quality of the loans made and sold at places like Countrywide had no relation to what the actual quality was. Unless Upstart is purposely misrepresenting things I don’t think this is a big risk. And I think the odds of that are reduced here as well after the COVID-19 stress test.

But where things fall apart in the writeup for me is valuation. To get to his valuation assumption, the author assumes LendingClub is fairly valued in terms of price/sales, that Upstart and LendingClub are largely identical companies except for growth rate, and then corrects for Upstart’s twice as high 2021 projected growth rate by giving Upstart a 100% premium over LendingClub’s price/sales valuation.

But that only takes into account 2021. Prior to 2021, LendingClub was growing revenue in the mid teens percentages, then had revenue drop by 59% in 2020. The 64% growth rate projected in 2021 just gets LendingClub back to 2/3rd’s of where it was in 2019, whereas Upstart’s 157% growth is on top of a 42% growth rate in 2020. Also, unlike Upstart, LendingClub is not profitable and is not projecting to be profitable in 2021.

LendingClub Net Revenue (millions)


**2015    2016    2017    2018    2019    2020**
429.9   500.8   574.5   694.8   758.6   314.7
----------------------------------------------
y.o.y   16.5%   14.7%   20.9%   9.1%    -58.5%

He might be right that Upstart is over-valued, but his means for arriving at that assumption is lazy. Upstart and LendingClub should be valued completely independently from each other, based on growth rates and margins specific to each company. The piece started out strong in the reporting of facts section but fizzled out in the analysis section (the most important part).

Mike

20 Likes

I agree with junomean2, the conclusion being drawn in their presentation is really not robust at all, with very little thought put into it - particularly the argument that LendingClub is Upstart’s best comparable peer, and that therefore, Upstart is overvalued.

There are many problems priced into LendingClub’s stock not mentioned at all in the report, which are incredibly important, and thus make LC a terrible peer to Upstart.

  1. LendingClub has a history of loan fraud/misrepresentation. Its initial founder left the company in 2016 as a result. https://www.bloomberg.com/news/features/2016-08-18/how-lendi…
    LendingClub also recently had to settle with the FTC for deceiving its consumers and double charging.
    Upstart remains founder-led and fully transparent, proactively works with regulatory bodies and possesses the only No-Action-Letter from the CFPB.

  2. LendingClub clearly has inferior AI underwriting than Upstart’s AI.
    Its loan performance is worse than Upstart’s, despite having much higher weighted average FICO across its trusts.
    Even looking all the way back to 2017, Upstart’s 2017 trusts had wtd avg FICO 687 while LendingClub 2017 trusts had wtd avg FICO 692.
    Yet UPST’s 2017 trusts produced a cumulative net loss rate (CNL) less than 10% while LendingClub’s 2017 trusts had CNL over 14%. https://i.imgur.com/4sIm1Lb.png
    LendingClub, in fact, has been forced to increase its wtd avg FICO over the years to reduce its losses. Most recently, their trust had wtd avg FICO at 713 for 2020! While Upstart trusts have wtd avg FICO down to 674 for last year - and its loss rates still continue to be lower than expected for its “traditionally measured” riskier tier of borrowers.

Also: Their CEO stated in earnings call in 2020 “Higher income, higher FICO, lower payment to income ratios and importantly, our new loans are heavily focused on our existing three million members, those who have demonstrated successful past payment history with LendingClub. This is because loans to existing members have exhibited significantly lower losses than loans for new members with similar credit profiles and they also come at a much lower cost of acquisition.”
LendingClub is being forced to pivot to safe, known, low risk, repeat borrowers as they don’t have truly superior AI underwriting tech.

  1. LendingClub’s conversion rate is far lower than Upstart’s 22% rate and thus higher borrower acquisition costs. Upstart has increasing conversion rates due to its superior automation, AI marketing models and frictionless tech/interface.
    (LendingClub doesn’t share conversion rate info publicly. “LendingClub’s application-to-loan conversion rate decreased from an average of 6.2% in June 2017 to April 2018 to an average of 5.6% in June and July 2018.” This information is only known due to FTC charges against the company on deceiving its consumers. https://casetext.com/case/fed-trade-commn-v-lendingclub-corp…

  2. LendingClub is not profitable and continues to produce losses. LendingClub’s total loan origination amount is also expected to be lower than Upstart’s loan origination guidance for 2021.
    In fact, LendingClub was initially created for P2P lending. It has been forced to shut this down in 2020 to pivot towards becoming a bank itself.
    Upstart, instead, is an already profitable, fast growing tech company that is not looking to be a bank or compete with banks, but instead partner with them.

26 Likes

makes me think Yield Fanatic has a short position.

Why else would someone put together such a nice looking and thorough presentation then distribute it through Twitter if they thought it was a bad investment?

7 Likes

Does anybody know if Upstart has exposure to China? It’s not mentioned in this research report and I looked in the companies 10-K and 10-Q reports and could not find ‘China’ mentioned once. It is selling off harder today than pretty much any stock on my watchlist besides SE (even the Chinese ones).

Hard to tell if it may be just due to the run-up it had yesterday… all pops have been immediately sold off since the IPO lockup release.

Bnh

Does anybody know if Upstart has exposure to China?

I’ve looked at their financial statements and presentations trying to find any mention of international exposure and have found no mention of it. My assumption is that Upstart for the time being is U.S.-based only. Please guys correct me if I’m wrong here.

Given the heavily-regulated nature of the industry, expanding into other countries, if that is an eventual goal for the company, would take more time than for a non-financial.

Mike

3 Likes

junomean,

Seems there is some kind of loan purchase obligation. I could find the following in their most recent 10-Q:
“11. Commitments and Contingencies
Loan Purchase Obligation
Under the Company’s loan agreements with certain bank partners, the banks retain ownership of the loans facilitated through Upstart’s platform for three days or longer (the “holding period”) after origination, as required under the respective agreements. The Company has committed to purchase the loans at the unpaid principal balance, plus accrued interest, at the conclusion of the required holding period. As of December 31, 2020 and March 31, 2021, the total loan purchase commitment
included outstanding principal balance of $39.3 million and $82.4 million, respectively.”

The majority of their loans are not retained by their bank partners. From 2020 10-K:
“Loans issued through our platform can be retained by our originating bank partners, distributed to our broad base of institutional investors and buyers that invest in Upstart-powered loans or funded by Upstart’s balance sheet. In the year ended December 31, 2020, 21% of the loans funded through our platform were retained by the originating bank and 77% of loans were purchased by institutional investors through our loan funding programs.”
The fact that Upstart is relying on the securitization market and what it’s willing to accept is a risk I didn’t fully appreciate.

Hephaestus (long Upstart, but only 5% position)

6 Likes