Power To The Elbow

Upstart (UPST) uses AI powered applications to help personal loan underwriters better rate the credit worthiness of borrowers. They claim to approve more loans at lower interest rates without the paperwork that is usually required for a standard loan. I am having 2007 flashbacks. Yikes. 

People have been taking out loans longer than we have had the technology to record them. We are constantly finding ways to justify giving out loans to people who will have trouble paying them back. Why? My guess is that every 15 years or so a new crop of lenders come along who want to innovate their way to higher profits. If they can somehow loan to subprime borrowers and have them pay them back, they will truly become billionaires. So far, this strategy has always blown up. 

Upstart is trying with AI and different approval criteria. They  are also bundling them and selling them to institutional investors who are shouldering the weight of whether Upstart’s AI algorithms are actually good at finding good borrowers. Upstart is thus incentivised to hand out as many loans as they can and then sell off regardless of quality. As long as investors are willing to buy their loans, they will do great. Or in other words, AI powered, 2007 subprime mortgage style, personal loans. 

Has anyone seen an AI do anything well the first million times it tries? NO! AI learns from its mistakes and then it tries again until it makes a new mistake. It does this hundreds of millions of times and eventually “learns” what the right way is. This AI is learning from past loan data. The great question is, what data on personal loans has this AI been trained on? I hope it was from the 1980s, the last time interest rates and inflation were at these levels. Otherwise we are staring straight into the next long term capital.

Is it going up or down:

There are a number of factors that say this company will not make money and is going lower from its current price of $35.9:

• There is a consensus price target of 20.21 from 17 different analysts. 
• There has been insider selling. 
• Topline revenue is falling. 
• Company is not profitable now when most borrowers are still paying back loans. It has operating profit of -30% for the last 3 months but over -50% for the year to date.
• Share count is growing. 
• 72% of sales are from just 3 customers.
• Upstart has fair value adjustments in the third quarter of more than -40$ million. They have written a lot of bad loans already. 
• Upstart is required to repurchase their loans from their institutional customers if the loans do not perform. YIKES!
• Loans issued between 2021 and mid 2022 have underperformed investor target expectations.

Things that will make it go up:

• It uses AI and that is a popular new thing to invest in.
• 38% of shares outstanding are already shorted. If this goes down, there are shorts who will buy to close out their positions.
• $222.1 million remains available for future purchases of our common stock under the share repurchase program.

I am interested in selling this weeks 31 put for $.24. Using past trading data, it has a 15% chance of landing in the money. In order to do it, the stock price would have to fall over 13% in the next 4 days. 

If the stock fell to $31.5 I would roll my position out one more week. I would buy back my position at 31 for an estimated $1.12 and sell next weeks 29 for a similar price. I would continue this pattern until my position expires out of the money.

There are many many people out there that don’t like this stock. I am one of them. That being said, between all the shorts who will have to cover their positions and the companies repurchase program, I think there will be enough buyers to stop a free fall in the stock price long enough for me to successfully roll my position up and out. It is important that I am out before the next earnings release.

What would you do? Where am I going wrong?