Facebook is Still a Bad Stock
A quantitative follow-up to my initial piece from January 10th
New analysis is posted every Monday at 10 am ET. Subscribe to Theta Thoughts to get posts that give you a premium for your time directly in your inbox!
Disclaimer: I hold no position in Facebook/Meta. This and all posts on Theta Thoughts are my thoughts alone and should not be constituted as investment advice. Please read my comments at the end of this piece for details on coming changes.
Note: The company “Meta Platforms” will be called “Facebook” in this piece for simplicity's sake.
Acronyms: DAU = Daily Active User | MAU = Monthly Active User | ARPU = Average Revenue Per User | MAP = Monthly Active Person | ARPP = Average Revenue Per Person
Facebook geographic segments: US & Canada, Europe, Asia-Pacific, and Rest of World
Despite Facebook’s stock dropping 31.44% since I wrote my bear thesis on the company during the weekend of January 8-10, I believe the stock is still bad. The thesis I wrote about has only become more solidified, and after building a model I believe Facebook’s stock could have the potential to fall an additional 19% from its February 8th closing price of $220.18. The following are a list of key reasons that drive this conviction
The company is losing both Facebook DAUs and engagement across all apps in its most valuable geographic segment
The company is failing to grow at required levels in Asia-Pacific and has seen negative DAU growth in Rest of World (MAUs have been negative in Rest of World for two consecutive quarters)
The company is actively losing young users to TikTok, Snapchat, and other competitors
The push into Reels directly harms ARPU/ARPP due to the very poor monetization ability of short-form video and hurts Facebook’s strategy of trying to upcycle young users from Instagram to the Facebook app
Apple privacy changes are having a profound impact, causing a forecast revenue loss of $10 billion in the upcoming fiscal year. The loss in targeting is also having an impact on ARPU/ARPP, with the company’s guidance implying year-over-year advertising ARPU/ARPP could be flat or negative
The core business is more capital intensive than previously thought with no sign of Capex lightening in view
Reality Labs is losing more money than expected and at an accelerating pace, with 2022 operating losses likely over $15 billion. There is also no clear vision of what will materialize from these losses that could benefit shareholders
The churn rate across all apps is much higher than anticipated and growing, with Q4 churn across the family of apps at over 5% according to my estimates
I already wrote about many of these issues from the Asia-Pacific and Rest of World user growth concerns to Apple privacy in my piece last month. You can read about my thoughts about those in my initial piece. For now, I want to discuss the additional worries I didn’t initially write about.
The Churn Rate
Last month people were posting data of a surge in Q4 downloads in Facebook’s apps, saying this showed strength in the business. What those people and Wall Street did not consider was the increasing churn rate the company is facing as it both reaches saturation and loses retention.
According to data from AppFigures, Facebook’s apps saw 393 million downloads in the fourth quarter. Note that I don’t include Messenger because of the requirement to have a Facebook account to use it. You can see the breakdown by app in the image below
That’s a lot of downloads, notably around 80% of which were on Android devices. Despite this large download number, monthly active people across all apps grew by just 10 million in the fourth quarter.
Facebook doesn’t give the data for you to calculate an exact churn rate, but it appears that from a mix of failing to retain new downloads and existing user churn Facebook lost many users in the fourth quarter. From their monthly active people across the family of apps, that would be a quarterly churn rate of 10.5% assuming each download was a different person.
The actual churn is lower than this. People can download two or three of those apps at once, but it means the minimum churn assuming every person downloaded all three apps in the quarter is 3.5% or 14.75% annualized.
I’m confident in estimating that fourth-quarter churn was over 5%. Annualized that would be a churn greater than 21% across the family of apps. The churn rate has been growing over the past year, and the current trend is not sustainable for the company. I do anticipate increased spend to limit this as a result. Whether this is through Reels, more notifications, something new I can’t think of yet, I’m not sure.
ARPU Compression Race
As I’ve continued to dig through data and view shifts in Facebook’s initiatives, it’s become increasingly clear that they are in a race to retain as much engagement as they can. In my opinion, management is willing to retain engagement even if it comes at the cost of materially compressing their ARPU/ARPP.
Reels is an attempt to copy TikTok. The problem is that the media format which both these platforms are, short-form video, is extremely hard to monetize. Monetization of this media format is so difficult that content creators can generate more income from just one view on YouTube than they could from over 75 views on TikTok or Reels. In terms of how much money Facebook makes as a company, the comparison is not much better.
The problem for Facebook as a company is that user engagement has and is continuing to shift away from their legacy platforms to things like short-form video. A consequence of this is that to limit user retention from fading more than it already is, the company must rapidly pursue this media format.
The result of the shift is that ARPU/ARPP can be materially harmed due to the lack of monetization ability on Reels when compared to Instagram itself. I believe the data already shows this happening as the ARPP growth rate in the fourth quarter was just 10.32%, a significant deceleration from the 14.92% growth rate observed in 2020 despite far higher spending from advertisers.
My Model and Price Target
To preface, this section, like the rest of this piece, is only my opinion and should not be considered as investment advice. I’m merely sharing my thoughts based on my work. That work has led me to the opinion that Facebook’s stock could decline an additional 19%. Below is the base case model I made.
There are a few points I want to note of assumptions within this model
This base case assumes MAPs (Monthly Active People) rise despite my doubts about this.
ARPP = Average Revenue Per Person. I’m using this to capture all users on every app to paint the clearest picture possible instead of only using the shrinking Facebook count
Gives Reality Labs faster sales growth and smaller losses than the street assumes. This is merely an act of conservation as my focus was on building the ARPP/MAPs side of the model
Operating margins in 2022 for the core business are 30.72%, materially declining as salaries and Capex rise substantially
Buybacks continue. I estimate 4.43% of shares are bought back in this model incrementally through the year with a higher Q1 amount after the stock decline
Within this model, I also projected ARPP for 2022 in the following image
I project ARPP declines in the first two quarters of the year, with the annualized ARPP projected to grow at just 1.11% in my model. This is due to the loss of targeting, chasing lower ARPP/ARPU media formats, higher user churn, and lower user engagement.
In the end, I project that the core business will see sales growth of just 3.12% to $119.26 billion in 2022 and that the whole company will grow sales at 4.44% to $123.16 billion for the year. This is in sharp contrast to the sell-side which currently predicts that the core business and whole company will grow at 11.93% and 13.12% respectively.
While I’m generous with Reality Labs, my thesis about the core business means it does not get the same treatment. I expect operating costs to outpace revenue in the core business substantially and that EBIT for it will drop by 6.58% to $53.18 billion. With accelerating losses at Reality Labs, I also expect the company as a whole to see a 19.05% decline in EBIT for 2022. I keep the tax rate the same and make almost no adjustments to non-operating income, so I anticipate net income to drop by a similar 18.90% to $31.93 billion. After factoring in buybacks, I currently expect that EPS for 2022 will contract by 16.23% to $11.79. This is over 6% below the average Wall Street estimate and includes conservatism on losses for Reality Labs of $2.4 billion. If I matched their expectations this would bring my EPS down to $11.
That’s how I got my earnings expectations, but what about the valuation side? Currently, Facebook trades at a trailing P/E ratio of 15.65x. With the struggles facing the business and the speculative nature of Reality Labs, I do not see a reason for this multiple to expand. I believe it could contract further, but for a conservative argument in making a price target, I hold the 2022 P/E ratio in my model at 15x.
The resulting price target from an EPS of $11.79 and a P/E multiple of 15x is $176.78. This would represent an additional downside of 19.71% from the closing price on February 8th for my base case.
The key risks I see in this are the following
The company materially reduces churn and resumes faster user growth
Reels becomes much more viable as a competitor to TikTok, and short-form video becomes more monetizable than I currently expect
What Reality Labs may do in the future to provide shareholder value becomes clearer to investors
The core business is less capital intensive than I currently anticipate
I could very easily be wrong in my expectations, but a further decline is what I’m currently seeing. Thank you for reading.
Normally I publish a research piece only on Mondays, but after talking to a professor today and evaluating my writing style and future desires, I thought it important to put this out despite the irregular time. When I leave college, I want to be a stock picker and eventual portfolio manager, not an investigative journalist. This means my writing will change from being just narrative-oriented to including more numbers and potential returns. To be clear, none of this will be investment advice, but if I want to get hired as a stock-picking analyst when I graduate, I need to work on this skill further.
I also got a good reminder that I shouldn’t be writing these to readers as if they’re an audience, but investors. That’s what you are at the end of the day. It’s going to be an adjustment, but I hope this is a step in the right direction. Thank you for your support and see you after the Super Bowl. Go Bengals!
Facebook is Still a Bad Stock
Adding more assumptions doesn't make something "analysis". Muted.
https://cdn.discordapp.com/attachments/764703716217323540/942265869264900106/vocodes_eddae000-0d03-4ad8-95b9-1a6bf2a9cabf.wav