Hi Fools,
Firstly, it has been a privilege to be able to learn from everyone on this board. Thankyou Saul and board members. I love how it is run - disciplined, topical and for our collective learning and benefit.
I was introduced to this board from someone who has been following Saul for over 10 years. Since being introduced to the board (only a month or so ago), I have read Sauls’s Knowledgebase end-to-end 3 times. It was literally like plugging myself into the matrix… all of a sudden: “I know Kung Fu!”…
Over the past month or so, I’ve had some time off work – a bit of rest. It has been nice. I have used much of my time researching many of the companies mentioned on this board. I have often spent 10 hours a day undertaking research and due diligence to inform my investment decisions.
As I am relatively new to these companies, I feel I can bring fresh eyes. A non-bias and objective perspective. Throughout my research only two companies have REALLY stood out to me. And of those two, UPST is by far what I consider to have the most opportunity over both the short and long-term.
I will disclose at this point that I am LONG UPST. I have bought in at $116, $125, $133, $136, $163 & $168. I’m likely to buy a little bit more, as I believe it has a long way to go yet.
I note from Saul’s most recent post, that GauchoRico mentioned back in April that he might lose 50% or make 2000% on UPST. I’m inclined to believe that the latter is entirely feasible. In fact, I mentioned to the person that introduced me to this board, when UPST was at $116 (before Q2 2021 earnings report) that from the research I have done, I believed that UPST could be the next AfterPay (ASX: $1.00 [2016] to $99.00 [2021], recently acquired by SQ). Needless to say, I have invested (responsibly) in UPST. I made this decision on my own, but was alerted to Upstart by this board – so thankyou to Saul, GauchoRico, Bear, Ethan (and to the person that intriduced me to this board - you know who you are) and to whomever stuck their neck out suggesting this company may be onto something. You have been both insightful and brave.
UPST – What does the future really hold?
Let me open by saying that I believe that Upstart will transform how banks assess creditworthiness (and credit risk) forever, as the iPhone has forever transformed the way in which we communicate.
Let’s just think about that for a moment… we could be on the edge of something just so ridiculously huge, its unfathomable. The early success of Upstart suggests that they are on their way from being “Good, to Great” to quote Jim Collins. Or maybe even Great, to Incredible (but I don’t want to get too emotional here – lets stay rational and keep our thinking divergent).
This ‘analytical thought-piece’ is mostly focused on Upstarts moat and TAM. I discuss what I believe to be some of their future opportunities, and I also touch on some risks and threats to Upstarts business. Please add to them to build our collective knowledge. I haven’t undertaken any financial modelling or statistical analysis. I will leave that to the experts on this board. If anything I have written changes your modelling assumptions and outputs, please share this with us.
FICO vs AI
The FICO Score® was developed in 1956 (55 years ago) by Fair, Isaac and Company (FICO). They developed an algorithm, which is still used today, and is the current standard to assess someone’s creditworthiness for a loan. Its an old system and it is quite simplistic in modern terms. A FICO was designed to be calculated by a human without a computer, but it always took quite a lot of work to generate an output. Fair, Isaac and Company registered a trademark for the FICO Score® with the U.S. Patent and Trademark Office all the way back in the late 1950’s, and we’ve been using it right up until today.
There is also another system called VantageScore which is a product produced by Equifax, TransUnion and Experian. The criteria is the same as a FICO, but only requires one month of credit history, in comparison to FICO’s requirements for at least 6 months of credit history.
And then along comes Upstart shaking everything….well, up. Upstart have now proved with a high degree of confidence, evidenced by significant volumes of quality data, that that the FICO Score® and the VantageScore are no longer the benchmark for creditworthiness assessment. The new benchmark is Upstarts AI algorithm, the new FICO – possibly, the new industry standard. Its notable attributes include that it is almost fully automated, almost instantaneous, and far more robust than FICO or Vantage.
What is evident to us already is that banks who partner with Upstart by embedding their artificial intelligence (AI) technology into their creditworthiness assessment process, are writing more loans, turning greater volume, suffering lower delinquencies and ultimately becoming more profitable. If those are the facts, then why would a bank not adopt this technology?
Total Addressable Market (TAM)
For now, let’s just talk about the US lending market TAM. These numbers are approximate and are from various sources, some of which are more reliable than others – so lets just treat these numbers as proxy’s to demonstrate market scale and proportionality.
- Personal Loans Market: $84 billion pa
- Student Loan Market: $ $446.2 billion pa (only $86 billion is private lending)
- Credit Card Market: $XXXX pa [no reliable data sourced]
- Auto Loan Market: $635 billion pa
- Mortgage Market: $2.35 trillion pa (2019 – new loans and refinancing [68% of consumer debt])
- Total US Credit Market: $4.2 trillion pa (new debt drawn pa)
- Outstanding Credit Card Debt: $756 billion (Q3 2020)
- Outstanding Auto Debt: $1.3 trillion (2019)
- Outstanding Mortgage Debt: $16.01 trillion (2019)
NOTE: Overall consumer debt is growing at 4% pa since 2015
For items 7, 8 and 9: I will discuss why outstanding debt is an important figure in the opportunities section. You’ll see why I believe these numbers are relevant… but they have NEVER been addressed in any Upstart earnings report because I don’t think they have considered how this market can be tapped. I will explain why I believe outstanding loans figures are highly relevant and could be monetized by Upstart (very exciting….).
Personal Loans and Auto Lending. The $719 Billion personal and auto loan market has only just been tapped by Upstart. Upstart only has five lenders using their software, of which only a couple of them would have made any sizable contributions to Q2 2021 revenue figures. The upside of this is that the new banking partners will be currently writing loans that are yet to be realized in the next quarter. It is evident that Upstart are very successfully moving toward the ‘enablement business model’ as a fintech. The most amazing part is that Upstart are already highly profitable, and this is with only five lenders active (of which only 2 would have made a meaningful contribution to Q2 2021 revenue) whilst only having only processed 0.9% of the personal and auto loan TAM volume. Incredible.
Mortgage Market. As we all know, personal loans and auto loans are small-fry in comparison to the mortgage market. Mortgages account for $2.35 trillion pa (2019 – both new loans and refinancing). Upstart processed a very impressive $6.682 billion of loans over the past 12 months, however none of this volume is from the mortgage market. If they can crack the egg with say, Bank of America, then this horse is going to bolt….faster than it is already. A very, very long runway ahead indeed.
OPPORTUNITIES
Almost instant access to a banks full lending volume and therefore instantaneous 100% revenue stream. The data just released by Upstart in the Q2 2021 earnings report was off the charts. The data have now demonstrated that Upstart can almost monetize a banks entire lending volume almost instantaneously. This is almost unparalleled in any other business – no ramp up (just initial system customization and integration), with no further infrastructure or logistics investment. Upstart generates instantaneous revenue which is reproduced by others (YoY) who are incentivized to maximize their lending volumes. Its almost too good to be true.
Whilst not the same industry, Amazon grew at 300%pa over their first few years. Google doubled revenues year on year in 2004 and 2005. But never have we seen the likes of this until SaaS and the Cloud. Coinbase went 12-fold revenue increase YoY a couple of days ago.
Revenue generation model. The revenue generation model is almost flawless, mainly because the volumes are continuous year over year, the product is so sticky, Upstart clients don’t pay for it (the loan applicant pays for it as part of their loan costs), banks make more money when they use Upstarts product and the volumes are simply MASSIVE.
Volumes from current partners are only the tip of the iceberg. At present only 5 banks have partnered with Upstart (as at Q2 2021). However this has only been very recently and it is unlikely that all their new banking partners have contributed a Q2’s reported lending volume. Telhio Credit Union only announced the partnership with Upstart in July 2021. NAFCU provided public endorsement in June 2021. KEMBA in March 2021. Oriental Bank in January 2021. So there is plenty of volume that has not been realized yet from these existing clients, and new clients will also come on board through Q3. Further, and more importantly NAFCU’s endorsement is big. Very big. They represent over 180 Credit Unions across the US (Upstarts Q2 numbers only represent partial revenue generated from maybe 2 or 3 of the 5 credit unions they have signed). Even better is that NAFCU now also advertise Upstart on their website which is VERY exciting. Link here: Upstart | NSC | NAFCU
Focussing on a fintech pureplay – not being a lender. The sooner Upstart remove themselves as a competitor in the personal and auto loan market, the better. Now that Upstart have proven that their AI tool is superior to all existing credit assessment methods, they can simply focus on being the enabler for the entire lending market. Their moat is the AI tool itself, which will be very difficult for anyone else to replicate. Therefore, for at least 2-3 years, they should have a monopoly, and an ability to monopolize the entire market.
Imagine the simple pitch to the banks or credit union…. “lending origination conversion rates will rise from typically 11% to 24.4%, delinquency should decrease by 75%, data shows that 71% of loan originations can be fully automated, you can reduce your loan origination staffing costs, provide a better product to customers, and become more competitive and profiitable.” Almost a no brainer.
Undervalued stock, as most do not understand the potential. For some reason, unknown to me, this stock seems seriously undervalued. Given that the business model simply relies on software to make some calculations, and then takes a clip of the ticket, with little to no risk implication, and having an enormous TAM – why is the company not valued like a SaaS company would be? It would seem as though valuations by analysts consider Upstart as a bank, not a FinTech. When the cat is out of the bag, analysts may very seriously upgrade their valuations where the forward P/S looks far more aggressive. It would be good to have some further discussion about this on the board.
Partnering with mortgage lenders. The untapped $2.3 trillion + market is awaiting. Whilst Upstart may not be able to charge the same percentage fee that they do for personal and auto loans (or maybe they will), the volumes that can be instantly enabled with each new banking partner is enormous. If Upstart was able to monetize only 5% of that market, their turnover volume would increase by $100 billion pa (vs. current volume of $2.8 billion this past year).
A risk mitigation tool, and a fee for service. Whilst not yet considered by Upstart (as far as I am aware), the AI tool will not only also facilitate a bank to assess credit worthiness for new loans, but will also, just as importantly, be able to fully automate a credit worthiness checks on every single one of their existing customers who have outstanding loans. This will allow banks to assess where they have sub-prime exposure, and therefore give banks an opportunity to forward adjust their risk profile and mitigate extant risk – before the next market meltdown. Upstart may be able to monetize the entire outstanding loans volume when banks run credit checks on their existing borrowers. If every single outstanding loan had a credit worthiness check (which triggered the usual fee on the value of the outstanding loan) and each bank conducted a credit check on every single borrower every quarter as part of their risk governance processes – then Upstarts revenues would not only be pegged to new loan volume, but to outstanding loan volume as well! Given that outstanding auto loan debt is estimated to be $1.36 trillion (2020), this presents a huge (and at present, unconsidered) opportunity. Whilst this may trigger some minor additional costs for banks, in all likelihood, banks would see a huge return on this investment as they would further reduce their delinquency rates by identifying at risk loans before borrowers actually default. What a brilliant value-add. It could be marketed by Upstart as a completely new offering, and could monetize a revenue stream that is not currently valued.
Potential partnership with Visa & Mastercard? Even though credit card debt is issued by a bank, by ‘swimming upstream’ to the card issuers themselves, it may be possible for Upstart to embed their software with the card issuers for banks and customers to use when applying for a new credit card. So instead of Upstart having to approach each bank to establish a credit card lending credit assessment agreement, the banks and their customers could automatically utilise Upstart as it would be embedded into VISA or Masatercards application suite. This approach is complex, and ‘has hairs on it’, but it may present an new moat and subsequent new enormous global TAM.
Becoming the industry standard for creditworthiness and credit risk assessment. In the longer term Upstarts’ AI may become the accepted industry standard (the benchmark) upon which all credit worthiness is measured. All lenders, all banks, all credit unions (even loan sharks) would need Upstart to be not only competitive on the market, but to give assurance to their investors and debt financers that they are effectively managing their lending risk profile. As such, all banks would almost ‘have to’ adopt Upstarts technology to satisfy their risk committees, lenders and investors.
A better S&P / Moodys credit rating for banks? Banks may also get better credit ratings if they use Upstart due to decreased delinquency rates and use of AI is their credit worthiness and subsequent sub-prime exposure risk.
For lenders, it lowers their risk profiles, improves their volume of lending, reduces their staffing requirements (and therefore costs), improves their access to capital, reduced their cost of debt and subsequently improves margins and overall profit.
For the customer, this means better lending rates, an increased likelihood of securing finance, and an improved customer experience.
Overall a win-win-win outcome (customer, bank, Upstart).
RISKS / THREATS
I can’t see that many risks that pose any game-changing threats to the business (war-stoppers if you like), but here are some I have considered. Feel free to highlight any more to ensure we have considered our exposure risk. Given I’m sure most of us are fairly loaded up on UPST, the better we understand the risks, the better we can determine what level of exposure we are prepared to accept as the value of our stock holdings climb.
Time it takes to onboard each bank. Product customization is a great opportunity that is already being offered by Upstart, however what this demonstrates is that Upstart is not a simple plug-and-play software patch. Upstart have stated that clients can tweak the algorithms to adjust for their own business’s risk appetite. The onboarding time for a banks may be significant due to banks strict risk assessment processes, governance, fiduciary obligations, complex software integration, and updating process and procedures, etc. I would expect that this would take between 3-18 months to integrate Upstart, depending on the scale and sophistication of the lender. Further, it is likely that sophisticated lenders will implement an initial pilot program to assure their risk committees that Upstart brings the benefits as claimed. Therefore, full lending volumes may take 3 to 12 months (or more) to materialize into earnings reporting. If this assumption is true, then Q3 2021 may be another bumper quarter.
Is competing in the loan market detrimental to growth? If you go onto www.bankrate.com, you can still get an UPSTART branded loan, which directly competes with Lightstream, Upgrade, Best Egg, SoFi, PenFed, Payoff, Marcus by GS, Prosper, Avant, and Lending Club. Upstart now have an opportunity to remove themselves from the retail lending market and focus on the fintech pureplay to enable ALL of their current competitors with their AI tech smarts. Whilst they are competing, it may be more difficult for them to grow their market share within the $84 billion personal loan market. At present they may be viewed as a ‘frenemy’, therefore onboarding new lenders may be more challenging. In saying that, the data that Upstart are able to gather through their own lending, may still be beneficial, as it would allow them to continue to assess data, and continually improve the AI tool. A careful balance is no doubt being carefully considered by the executive team at Upstart. Hopefully Upstart are able to confidentially access and assess the data they gain from their banking partners to facilitate continuous improvement of their technology.
Finally, if Upstart can make this shift to becoming a pure fintech company (which their Q2 2021 earning call report suggested they may be working towards) this could drastically improve their market valuation. Historically (seemingly to me) Upstarts valuations has been based upon Upstart being a loan provider with a fintech edge. Now Upstart can look towards a valuation as a pure fintech with a banking moat, which may see a significant value upgrade and analysts setting a new target share price.
Market cycles. Revenue will be pegged directly to lending volumes. As such, the cyclical nature of lending volumes will be directly reflected in earnings. This could result in revenue volatility. The upside to this is that in times of stress, those banks that are using Upstart may seeing improved lending resilience and lower delinquency than their non-Upstart banking peers. Hence, new banking clients would soon want to enable their business with Upstarts technology.
Out-sourcing. There may be resistance from the big banks to outsource to Upstart. Banks may not want to become reliant on a single third-party for all of their lending. What happens if the two parties fall-out of favour with each other? What is the mitigation to such an event that Upstart simply ‘turn-off’ their AI software during a dispute? The fintech ecosystem is highly dynamic, where competitors are ironically reliant on each other for survival. The frenemy is becoming the norm in this complex environment. Sophisticated contracts will need to be developed to overcome these unique risks. I don’t know what the mitigation to this risk actually is, however what I do know is that the big banks that sign with Upstart, set to benefit from first mover advantage, which will improve their market share through the ability to write more loans.
Competition. I have not deep dived into this too much, but the only company I can find out there that has a similar tool is Zest AI, who are working with VyStar Credit union (one of the largest CU’s in the US). It seems Upstart have the jump on them, and their technology is far more mature. But what is does show, is that Upstart needs to make the absolute most of being the first mover in this new market. To date, Upstart are certainly succeeding on this front.
Ability to grow into overseas markets. Upstart relies on open data sources. The US is well known for its minimal levels of regulation, and therefore very good access to large swathes of individuals data. If Upstart try and replicate this in other countries, the ability to source the 1600+ data points to satisfy their AI algorithms will prove very difficult. In fact, the red tape in some countries will make it almost impossible for their AI model to even work. For now, the US is the perfect market to grow and establish their AI capability. Upstart may become so successful, that large banks actually reach out from all over the world requesting Upstart to come and help them build an AI product suitable for their country, laws, market and business model.
CONCLUSION
Given the extensive opportunities that still await the Upstart business, I have a very high level of confidence that Upstart will continue to succeed, and that they will very quickly capture and dominate their new market. Their short track record has been impressive, which improves the level of confidence I have in making this statement.
Their AI product seems to be performing to a very high standard and Upstart have statistical evidence to prove that this is the case. The product has now made the breakthrough into mainstream credit unions (having recently being endorsed by the National Association of Federally Insured Credit Unions - NAFCU) which will provide instantaneous momentum as credit unions and banks hustle for first mover advantage. It is worth noting that NAFCU have over 180 credit unions within their membership!
There is a long runway ahead given Upstarts enormous TAM. The personal loans, auto loans and mortgage markets are in the trillions, of which Upstart only captured a volume of approximately c$3 billion to date.
Finally, the opportunity for Upstart to tap into volume for existing outstanding loans is outrageous. If Upstart can demonstrate to banks that using Upstarts AI to assess the creditworthiness of their current customers with outstanding loans will help improve banks to manage their current outstanding debt risk. To understand ho much of an opportunity this is, the outstanding auto loans TAM is $1.3 trillion. The mortgage outstanding debt is $16.01 trillion (2019).
Whilst Upstart are still to to breach the mortgage market, or any of the ‘big banks’, Upstart are quickly carving out a very profitable and expansive market within the credit unions which are focussed on personal and auto lending. Given that only 5 of the 180+ credit Unions have signed up to Upstart so far, Upstarts current focus market is still extraordinarily deep.
The share price as I finish this post is $183.52. Whilst the share price has risen very quickly since my first BUY at $116 (justifiably based on earning results achieved and forecast) I think we will look back in two years time and say…“I wish I bought more at $183…”.
I see plenty more upside in Upstart.
LongMalibu
PLEASE NOTE: I am not a professional investor. This is only an opinion piece and you should conduct your own research and validate all of my work before making any decision to invest in UPST.