Description

C3.ai is a stock pump scheme masquerading as a software company. I believe it will eventually be a zero - although the company’s $700M cash balance (from their 2020 IPO; burned $160M LTM) will result in a slow bleed rather than a sudden death. I have been short the stock on and off since January 2021 and began pulling together this report last week – I am posting it now because I think it’s a particularly attractive short entry following the latest “Generative AI” press release. The cost to borrow is 0.5% and I believe this is the best risk-adjusted way to short this January squeeze. 

C3.ai is a hugely unprofitable business with no product differentiation, huge customer concentration, extreme executive turnover, and a highly promotional 70 year old CEO. The company claims to be an “Enterprise AI Platform,” though it is essentially a consulting shop with 32% of revenue coming from a faltering relationship with Baker Hughes and 39% coming from their next two largest customers (one likely the Air Force). 

Sequential revenue growth is down each of the last two quarters, and cRPO / RPO are now negative YoY (though the company is transitioning to a “consumption-based” model).  

The company came public in December 2020 and soared to a stratospheric $20B, ~100x revenue valuation despite 6% revenue growth in the quarter ending October 2020. The stock is down materially since the IPO, though CEO Tom Siebel has sold more than $600M worth of stock. The company has approximately 140M FD shares outstanding, which equals an Enterprise Value of ~$1.7B at $20 / share. This Enterprise Value is up 3x from ~$500M in December. 

The stock is up ~70% YTD due to AI hype and the company’s recent (nonsensical) press release about providing a Generative AI solution. There is obviously some meme / squeeze risk here, but I have personally never felt more certain about the long-term negative outlook of a business. 

There are three key tenets of this short thesis:

1. Company History: Highlights the non-seriousness of this business

2. Executive Turnover: Extreme sales, leadership, and CFO turnover

3. Business Slowdown: To the extent there is real economic activity at C3, it is worsening quite rapidly

I will discuss the business results last since I believe it is less important than the cultural shenanigans at this company. 

I would also direct you to the Spruce Point short report (https://www.sprucepointcap.com/c3-ai-inc/) from February 2022 which goes into greater detail about specific issues. 

Company History

Tom Siebel is a legitimate (historical) technology executive. He was an early executive at Oracle in the 1980s and founded Siebel Systems, a CRM system he sold back to Oracle for $6B in 2005. In 2009, he and several friends / close colleagues (including current board member Pat House and current CTO Ed Abbo) founded C3 Energy to take advantage of opportunities around clean energy (it was not clear what C3 Energy was supposed to do). 

Siebel is known as being a tough boss and pugnacious figure. He survived a horrendous elephant attack in 2009 while on safari and drew ire for his demolition of historic buildings near his Montana ranch as well as donations to recent Canadian trucker protests. He is closely tied to Stanford and leading conservative figures (Condi Rice is a board member, and the company has an advisory board with several prominent retired Generals). The board could be described as a “Celebrity Board.” 

In 2016, the company changed its name to C3.IOT and positioned itself as a leader in IOT, particularly focused on manufacturing and oil and gas. In 2019, the company again changed its name to C3.AI. The company claims to now have multiple product lines, including a “smart” CRM, C3.ai ESG, and C3.ai Law Enforcement. Some of C3’s marketing indicates that it is a market-leading CRM. 

In March 2022, Siebel described what the company does in an unintelligible way: 

“Now this can be a bit difficult to get your mind around, but if you think of a single integrated AI software solution that provides, in one package, the union of the functionality offered by all of the AI software companies that you are aware of, say, Palantir, Dataiku, DataRobot, Databricks, DataStax, Alteryx, TensorFlow, et cetera. That's what we offer to the market. How do we do it? The secret sauce of C3 AI is this concept of a model-driven architecture. Our approach dramatically simplifies complex enterprise application problems so that our customers can rapidly design, develop, provision, operate industrial and commercial scale AI solutions.”

And on January 31, the company put out a press release announcing its new Generative AI Suite (https://www.businesswire.com/news/home/20230131005526/en/C3-AI-Announces-Launch-of-C3-Generative-AI-Product-Suite):

“The C3 Generative AI Product Suite integrates the latest AI capabilities from organizations such as Open AI, Google, and academia, and the most advanced models, such as ChatGPT and GPT-3 into C3 AI’s enterprise AI products.”

The company took several other spurious actions to pump the stock in the past, including a press release indicating a “$500M award for the Department of Defense” (this was merely an authorization for the DoD to spend that limit, and no business was won) and announcing a $100M buyback, which they stopped after $15M of purchases even though the stock then declined >50%. This is not a serious company. 

Executive Turnover

The company has approximately 950 employees, up from around 500 at the December 2020 IPO. They are on their 5th CFO since June 2019 in Juho Parkkinen who was most recently a mid-level (Senior Director) MongoDB accountant. The only other executives they have listed as leadership are Ed Abbo (CTO, former Siebel Systems, at C3 since 2009) and President Houman Behzadi (President, former Siebel Systems, at C3 since 2010). 

Spruce Point highlights 11 VP+ level sales execs that left in the second half of 2021. They lost at least another 18 VP+ GTM execs in 2022:

If you go through LinkedIn, it is hard to find any remaining senior GTM folks. 

Business Slowdown

~33% of the business comes from Baker Hughes. In 2019, Baker Hughes signed a 3 year, $320M deal with C3 where Baker Hughes received 15% of C3 and a board seat. Since then, the agreement has been renegotiated in Baker Hughes’ favor twice (longer duration, lower near-term spend, ability to resell commitment). Baker Hughes CEO Lorenzo Simonelli resigned from the board in late 2021 and BH has reduced its stake in the company down to ~7%. 

The service delivered to Baker Hughes can most generously be described as a custom-built predictive maintenance solution. This is essentially a services business that claims to be selling software, but with well under 100 customers it’s clear that they don’t really create true software products. 

In the quarter ending October 31, C3 grew revenue 7% YoY and -4% QoQ to $62.4M. Their guidance for the quarter ending January 31 implies -7% YoY growth - and they recently missed guidance. Leading indicators of growth, including cRPO and RPO, are both negative YoY. 

Non-GAAP Operating Loss has ranged from $15M - $22M over the last 7 quarters, but next quarter guidance is for a widening loss of $25M - $29M. Stock based comp in the most recent quarter was 80% of revenue - though, in their defense, dilution is slowing and that mostly reflects excessive compensation made in 2020 / 2021. 

Therefore, the current Enterprise Value is 6-7x NTM revenue with no growth, and potentially negative growth over time. At best, this should trade like a services business for 2-3x revenue (~50% equity downside). At worst, their high burn rates will chew away at their cash and the equity will eventually be zero. 

In sum, this is a shrinking business, with widening losses, with no real product, and with huge executive turnover and a history of brazen stock-pump-by-press-release. If you can tolerate some potential AI-related squeeze risk, this is a short that should pay tremendously well in 2023. 

Catalyst

AI bubble subsides

Earnings in March followed by insider sales