Meta: Investing for An Exponential Future
Deep Dive
META 0.00%↑ set off my contrarian alarm, so here is a deep dive on the company. Enjoy.
1.0 The New Flywheel 2.0 Stigma, Culture and Transitions
3.0 Framing the Metaverse 4.0 Financials 5.0 Conclusion
1.0 The New Flywheel
Meta is gearing up its Family of Apps for its second, perhaps far more explosive life, by creating a new flywheel that can supercharge it.
Since its IPO, Meta has been brewing a series of platforms and interconnecting them, if only rather timidly. After Q3 ´22, the platforms reach/serve just under 50% of the world´s population and yet, the company has now conferred a stagnant feel, as a series of factors get layered on top of the company´s ongoing investment spree. Over the last few years, the dominant narrative has become that TikTok (which is a CCP weapon) will displace the company´s crown jewel Instagram but meanwhile, Meta has been developing a new flywheel with enormous potential.
Other than the Metaverse, which merits a section on its own, the company has been investing heavily on two domains, which are bound to retro-feed as they scale out over the 3.7B user base and create very powerful network effects:
AI
Business Messaging
Today, 15% of the content viewed by users across Facebook and Instagram is recommended by AI and the messaging surface concerns mostly P2P interactions. Even at this seemingly harmless percentage, as a user it may be apparent to you that the recommendation algorithm is getting scarily addictive. You check into Instagram to check something and all of a sudden, you are on your tenth Reel about Corgi dogs. You may then finding yourself DMing that Reel to another friend, which is likely soon to be on a similar loop.
What happens when the % of content recommended by AI rises to say 50% and users are able to directly interact with brands via messaging? You end up with a worldwide machine that creates viral loops at a marginal cost and is then able to convert that traffic into commerce, in a highly granular and efficient manner. Something like the current Family of Apps, but supercharged, far more independent of the operating systems that support it and also somewhat damaging for the world´s mental health.
A sub-narrative that I have found in my research is that Meta is failing miserably at monetizing Whatsapp and its other messaging surfaces, but reviewing the company´s quarterly advances in depth since Q1 2020, I have found that it is actually doing quite well:
In Q1 2020, the company was just laying out the vision for the “click-to-messaging ads”. Per Q3 2022, the business is now at a $9B annualized run rate. From Q1 202 to Q3 2022, the company has been gradually getting more vocal about the performance of this business segment.
In Q3 2020, Reels was an infant and just getting rolled out to 50 additional countries. Per Q3 2022, there are now 140B Reel plays every day across Facebook and Instagram. There are also 1B Reel shares on Instagram daily.
In Q2 2022, Reels accounted for 20% of the time spent on Instagram, representing a 30% sequential increase from the previous quarter.
The minor caveat with Reels is that the company has not yet found a way to monetize them and in turn, it draws away traffic from other highly monetizing surfaces, like Stories. In Q3, this had a $(500)M yield as it did indeed absorb traffic. As the growth continues, I believe it will continue to hurt the financials. Nonetheless, per the aspects of the company that I explore in section 2.0, I believe that in the longer term it will not be an insurmountable problem for Meta to find a way to monetize Reels effectively.
“ … any time we’ve had a new format, like when we added Stories or even before that when we were primarily on desktop and we shifted to mobile feed, we sort of had this dynamic where we focused on increasing engagement and growing demand for the product.”
“ … Reels is actually making faster progress than we'd expected. We've now crossed $1B annual revenue run rate for Reels ads, and Reels also has a higher revenue run rate than Stories did at identical times post launch.” - - Mark Zuckerberg, CEO, Q2 2022 ER
As illustrated by the % of content that is recommended by AI, this is only just the beginning. Fundamentally, the company is aiming to monetize this new emerging surface by incentivizing merchants to point their ads at the messaging API, instead of charging for the latter and then building out a commerce infrastructure around the API, to facilitate the conversion of clients. This is a very agile way for the company to capture value and enables for the ad attribution to easier, in an increasingly privacy focused world (see ATT, iOS). In testing its new Advantage+ Shopping feature during Q3, the company found advertisers obtaining a 32% increase in ad RoI.
“Our approach is to work our way down the stack and build world class services at every layer of commerce – starting from discovery at the top of the stack all the way down to payments. Just like we want to be the best place for millions of creators to make a living, we also want to be the best place for businesses to grow as well.”- Mark Zuckerberg, CEO, Q2 2021 ER
If you have been reading my work, you will know that I am fond of companies that deploy initial electron management layers, to then add subsequent cash rich layers at a marginal cost. In this instance, Meta is a prime example, having first developed its initial platform that spans ~50% of the population (which many have criticized as un-inventive) and is now proceeding to enable their users to interact at higher dimensionalities. If the company keeps up the progress seen from 2020, it will not take long for it to show up in the numbers big time.
Finally and before moving on to dissect the company´s culture, which is a central part of the thesis, I want to address a number of considerations to the above:
I notice that the world is going up a dopamine ladder and that although it serves to connect people which does much good, it is also damaging people´s mental health. I try to invest in companies that are working on a future that I believe in and that I want to see for my offspring. Nonetheless, I suspect the way forward is more education on the matter, because the tech is greatly empowering for many humans too.
AI evolves non linearly and as such, it may prove harder than expected for Meta to catch up with TikTok. In effect, how TikTok is evolving is a black box to public market investors and so it is hard to benchmark the two companies. However, as I said above, I believe TikTok is actually a CCP weapon and it should be banned as soon as possible.
2.0 Stigma, Culture and Transitions
Meta has mastered transitions and its current situation is not too different from similarly life threatening ones in the near past.
I have until recently fallen into the trap of stigmatizing Meta. In researching the company for this write up, I have actually been very positively surprised with how people work inside the company and just how excellent the culture is, which stands in great contradiction to the immediate external perception. In summary, Meta is a kaizen machine and people generally love working there. In Zuckerberg´s words during Q2 2022 ER:
And culturally, we focus on moving and learning faster than everyone else. And I think that those are sustainable advantages.
But I think at the end of the day, what that really comes down to is just I try to push the company to be one that learns faster and just keeps iterating and moving faster than we did in the past and than others in the industry do. And I think if we can do that well, then we'll continue to succeed.
But I think the moment that we stop doing that, then we'll basically fall behind.
The company has now been through a series of transitions, in which the market questioned whether the company would survive. First, the shift from desktop to mobile and then the move from feeds to stories. Whilst many view companies as static entities, I view them as a collection of human beings working in a particular way, which ultimately tends to define the fate of a company. Meta´s ability to move forward and deal with challenges stems from its ability to iterate incessantly, as is the case with other world leading companies.
A deep dive through Meta´s Glassdoor valuations reveal the usual characteristics, that enable the company to be a kaizen machine. The valuations conform a very rich dataset and below follows a condensed representation of what I have picked up:
Information flows efficiently: the company has a flat hierarchy and feedback from lower ranking employees is well processed by management. Many employees have the sense, which may not actually reflect the reality, that the company is run bottom up.
There is a constructive attitude towards failure: employees are empowered right out of the gate, as evidenced by some accounts of interns being given relatively big responsibilities and are allowed to make mistakes and learn.
Employees find meaning in their work: people feel connected to the mission, there are a “small percentage of jerks” and feel like colleagues are generally nice and highly talented. Work is challenging and the organization moves fast enough for people to be engaged, with occasional complaints of work-life balance getting out of touch.
“And over the longer term, we're developing privacy-enhancing technologies in collaboration with others across the industry to help minimize the amount of personal information we process, while still allowing us to share relevant ads. Progress in these areas will take time and will be a focus for us throughout 2022 and beyond.” - mark Zuckerberg, CEO @ Q3 2021 ER
In effect, the company has to transition towards AI and privacy now and the above characteristics are more of a lifeboat than the market commonly believes. If you study tech companies across the board, it is actually quite rare to find kaizen machines in which people enjoy working because the balance is so hard to strike. People have to feel heard, connected and challenged enough to not lean back or get burned out and scaling this up to tens of thousands of employees is a fine art. The companies that do strike the balance tend to have a track record of pulling off one impossible feat after another.
On Nov 9 2022, the company layed off 11,000 employees, or 13% of its staff. Whilst I get the feeling that the market is looking at this from an OPEX perspective, I think the more important angles is the cultural one. Across Glassdoor, there are complaints of the experience of working at the company getting diluted, with increased scale. In effect, what this does is make employees feel less challenged and connected to the mission, upsetting the balance that enables kaizen. Per Zuckerberg´s words in Q2 2022, I believe this is the angle he has approached the layoffs from:
“This is a period that demands more intensity, and I expect us to get more done with fewer resources. We're currently going through the process of increasing the goals for many of our efforts. Previously challenging periods have been transformational for our company and helped us develop our next generation of leaders. I expect this period to be no different.”
The main issue that I see culturally for the company is the bifurcation that is currently happening: the company is splitting its resources between the next iteration of Family of Apps and its Metaverse efforts. Going forward, the main question is, where do the best engineers in the company want to work now? The nature of work also varies completely from one domain to the other, with work in the social media side being allegedly metric driven, whilst the work in Reality Labs seems to be more experimental. The company has issued no information on how this is being managed and so this remains a big black box for me in the thesis and perhaps, the most important aspect of it.
3.0 Framing the Metaverse
Reality Labs adds a predominant speculative component to the thesis, but perhaps it may suffice for the company to traction one high end device. At the right price, this can be an asymmetric bet.
Assuming that the flywheel exposed in section 1.0 works, the question for me is: how to frame Meta as an asymmetrical bet? The Metaverse investment can be a sunk cost, regardless of its potential going forward. Assuming that the flywheel exposed in section 1.0 works and continues to produce enough produce cash to sustain itself and the Metaverse endeavor, Meta becomes a truly asymmetric pick below $100B, the very rough combined value of Whatsapp and Instagram. I am being quite conservative and I am assuming that Facebook is toast, which in fact it is not quite, as you can see below.
Naturally, as the market cap moves in that direction the potential investment becomes gradually more asymmetric and it is hard to tell the exact value of Instagram and Whatsapp . This is more of an art than an exact science, so maybe at a market cap of $200B it looks asymmetric enough to me at some point. If this margin of safety seems unwarranted, however, consider history. Having listened to various natives talk about Reality Labs on podcasts, I get the sense that Meta is taking on a task that governments and entire industries have taken on in the past - that of developing a whole new platform, analogous to the phone for example.
If you review the history of any key piece of technology in today´s world, such as the train, each advance is marked by a tragic story. The first person to commercially apply a train was, according to Richard Rhodes (“Energy: A Human History”), a man who set up a circular track in London and enabled people to go on a joyride in exchange for a fee. He died riding his own (faulty) train, as did the inventor of the Segway and countless others, who perhaps did not die, but went through other kinds of misfortunes.
Meta´s Reality Labs is having to work on every component from the ground up and, apart from Meta’s pre-existing ability to iterate and get better at things, it seems to require a whole different set of engineering skills. History suggests that many will “perish” in circular tracks too. Whichever way you frame it, at least for outsiders, Reality Labs has a predominant speculative component and so a very reasonable entry price is a prudent requisite.
However, Meta does not have to instantiate the Metaverse fully for its efforts to bear fruit. I am particularly impressed with the Meta Quest Pro 2, which is a high end device that allows professionals to work with AR. It still looks somewhat clunky and the experience seems to be uncanny, but I can see how it could be a great success N iterations down the line . Once/if it gains traction with this device, the company then has a runway to incrementally add features via trial and error that bring it closer to the actual end goal. The upside here is infinite, as the company goes beyond the hardware and into platform services (identity) and apps.
My average work day consists in learning about many things, all through a screen. I would definitely pay the current price of $1500 to turn my work days into an immersive experience, in which I can explore things in a much more dynamic way. I believe this extends to many other creators across the world and that the applications in this sense are endless. One-to-many immersive sessions with your audience, live collaborations with other creators, all with objects flying around that humans can interact with. In one picture:
Apart from gaining traction with the actual device, Meta must expedite its transition towards privacy and I am looking forward to seeing actual advances in this area. It has a reputation for not caring about the latter and with this device, it can do great good but also a fair deal of evil. Via a headset that workers may grow to be dependent on, the company has a much more intimate hold on its customers than previously. Through it, Meta can hijack your brain far more easily than through a small screen half a meter away from you, although it currently does a great job. For all intents, it must come to actually respect its users integrally, which may very well not be in the company´s DNA.
4.0 Financials
“Our capital allocation philosophy over the long-term is to allocate a portion of the profits generated from the Family of Apps towards these future-focused areas while enabling a greater return of capital to shareholders.” - Mark Zuckerberg, CEO, Q3 2022 ER
Income Statement
It would make little sense for investors to attempt to frame Meta as an asymmetric bet, without management itself doing the same as it refers to Reality Labs. This is necessary to sustain the value of Family of Apps through time, which is what in the first place mitigates the risk of permanent loss of capital and as discussed, can make the investment non-speculative at the right price. In Q3 2022, the company directed the majority of its investments towards Family of Apps. During the quarter:
Family of Apps expenses were $18.1 billion, representing 82% of overall expenses.
Reality Labs expenses were $4B or 22% of Family of Apps expenses.
The company expects Reality Labs expenses to “increase meaningfully” again in 2023 and I do not have a problem with that, so long as the majority of the capital continues to be directed towards Family of Apps, so the asymmetry holds. Per Zuckerberg´s words quoted at the beginning of this section, I believe the company is currently thinking along these lines too. As discussed, the cultural bifurcation addressed in section 3.0 must be adequately managed for the capital allocation to be truly effective in the first place across the two business segments.
Additionally what is currently happening is that the following three things are being layered on top of each other:
A weakening ad spend together with FX headwinds, after what seemed like a permanent acceleration of digital business during the pandemic.
Meta´s current investment cycle, which is taking a toll on its bottom line and cash production.
The TikTok narrative.
What truly matters is whether the company, apart from pursuing the Metaverse bet responsibly, is able to iterate its Family of Apps onto the AI stage, thus successfully creating the flywheel I discussed in section 1.0. If it does, advertising profits will continue to climb for probably another decade and long term shareholders will be satisfied. The weakening ad spend is acting like a sort of mist that is not enabling the market to see this and as per usual, it is favoring short term profits over investments for the long run. All incremental CAPEX assignments that are prudently allocated to advancing the AI capabilities of the company are welcome on my end.
Developing AI capabilities is not only about software but also about hardware (you can learn more about this in my AMD deep dive). Meta is investing heavily on developing its own datacenters with the adequate “servers and networking equipment”. The company has issued no further details, but I know from my research in AMD that to really enable AI at scale, the networks that support it themselves must also become smart. The fact that the CAPEX is being largely driven by investments in such infrastructure encourage me to think that the company is committed to evolving its Family of Apps.
It is one thing to load an image to billions of people across the globe and another to fundamentally understand the content of the media itself. The latter requires exponentially more compute power and this efficiency is a key aspect of the operation, if not the most important one together with the actual user base that generates the content in the first place. I believe Meta is right to invest allocate capital heavily in this domain and I look forward to learning more about what the company is doing exactly here.
Cashflow Statement
Whilst revenue growth has slowed down recently coming of the pandemic, the track record over the last 10 years has been stellar and looking back, the key enabling feature seems to have been excellent capital allocation. The Whatsapp and Instagram acquisitions were particularly masterful and I remember that at the time, many thought that Zuckerberg was dumb to say the least, for paying $1B for Instagram. I believe that in fact, Zuckerberg is one of the better capital allocators that I have encountered and this makes a big difference going forward.
As the company is ramping up its investments, cash from operations is going on a round trip but it is still very much in the positive domain. The key going forward in this aspect is for the company to strike a balance between investing in the future of the business and generating enough cashflow to keep it healthy. Whilst there is nothing particularly important going on regarding cashflow from financing investing activities, cash from financing activities is marked by the company´s heavy stock repurchasing, which is a favorable tailwind for investors.
You may appreciate that since Q1 2019, the company ramped up its stock buybacks leading into the Q4 2021, trailing the general enthusiasm in the market. This somewhat correlates to the enthusiastic hiring that the company engaged in the same period, that it is now working to discard. This dynamic stands in slight contradiction to the company´s excellent capital allocation track record over the last decade and it seems that it perhaps got a bit carried away by the general euphoria. I am nonetheless glad to see Meta owning and correcting its mistakes, but here we see that the company is not infallible.
“Our pace of hiring slowed in the third quarter, consistent with our previously stated plans. We added 3,700 net new hires in Q3, down from our Q2 net additions of 5,700 despite Q3 typically being a seasonally stronger hiring period. We expect hiring to slow dramatically going forward.” - Dave Whener, CFO @ Q3 2022 ER
Free cashflow has been severely hit by the CAPEX ramp and Q3 2022 exhibits quite a sharp drop, even when factoring in the “$413 million impairment of certain operating leases as part of our ongoing work to align our office facilities footprint with our anticipated operating needs.” It is hard to asses this aspect, because as an outside observer I do not have granularity into what the minimum capital requirements are to drive the company´s key initiatives forward and thus, to what extent free cashflow generation really must be sacrificed in the short term. However, the drop from Q4 2021 to Q3 2022 exudes some degree of enthusiasm too.
Balance Sheet
The balance sheet is strong, ending Q3 2022 with $41.7B in cash and equivalents and $25.1B in LT debt, although the company is leveraging up. During the same period, the company completed am debt offering of $10B, notably magnifying the overall long term debt profile. This stands in contradiction with the company´s weakening cash generation.
It is also interesting to see that the company has never quite seen a short term debt spike in its public history. Short term debts tend to spike when a management team is caught by surprise and needs emergency funding of some sort. To me this signals, together with the obvious qualitative factors, that the company is well run in general. Below, short term debt in green and long term debt in orange.
The newly issued long term debt is callable (meaning that Meta can opt to pay it down any time it wishes to) and is structured as follows:
$2.7B with a coupon of 3.5%, due in 2027.
$3.0B with a coupon of 3.85%, due in 2032.
$2.75B with a coupon of 4.450%, due in 2052.
$1.5B with a coupon of 4.65%, due in 2062.
Per the cash from operations of the company, I believe it has the capacity to pay this debt off and specially so as the Family of Apps moves towards AI. However, I wonder whether this is the best time to emit such debt. For now it seems like higher rates are going to be part of the norm going forward, but perhaps the rate hike from 0% to 4%+ is in fact and infinity hike of unprecedented magnitude. When Volker went from 10% to 20%, he doubled the rate - going from 0% to 4% is a much bigger stretch and perhaps exponentially so. I believe, although I do not base my capital allocation decisions based on this thought, that at some point we are going to have to deal with deflation.
Perhaps in hindsight, this debt emission takes on the look of the stock buybacks in Q4 2021.
5.0 Conclusion
The things that have done the most good for humanity have also done the most harm. I am afraid that social media / Meta is one of them.
The purpose of having a research backlog is being able to take advantage of opportunities quickly when they arise. Meta is misunderstood, but not quite enough for me to make a move just yet. However, I see that management has clearly taken the decision to sacrifice short term profits for long term value creation and this often does not go well with the market. For this reason, I think there is a considerably probability that the asymmetric opportunity that I outline in this write up presents itself.
For the company to be in a much better position 10 years down the line, it must master AI. It has been on the task since early 2020 and thus far, the results are quite appealing. Further, if responsibly managed (and at the right price), the Metaverse bet can combine with its AI efforts to yield a monstrous company - perhaps one too powerful. Whilst many see the Metaverse as a distant and diffuse reality, I see it as starting with one hardware device that gains traction, much like the iPhone kicked off a whole new domain of the economy. Some degree of traction is all the company needs to then work its way towards its long term vision.
Having been a participant of the stigma myself for many years, I am very positively surprised by the way the company is run and how people work inside it. From an organizational stand point, it seems to be one of the better companies that I have come across and this, as is the case with many other leading corporations, is actually its key asset. So long as it remains well capitalized, agile and meaningful enough for its employees, the company will tend to do well in the future as it continues to iterate and get better everyday. So far, it seems that the company intends to remain a loyal kaizen apprentice.
Going forward, if the overall market´s perception of the company continues to worsen and the company gets nearer to traction on the Reality Labs side, this can turn into a very attractive pick for the long run. Finally, about the mental health considerations, I was watching this part of a Laird Hamilton interview, hosted by Rich Roll, in which they discuss just how addictive social media is but also how liberating and empowering it has been for people that chose to craft their own path in life. In effect, this brought me back to the major thing that I have learned from my history studies, which is that the things that have done the most good for humanity have also done the most harm - like money and religion.
I am afraid that social media is one of them and we are just getting started. If Meta plays its card well over the next few years, it can set itself up for an exponential future in which there is pretty much no limit to the value that they can add.
Until next time!
⚡ If you enjoyed the post, please feel free to share with friends, drop a like and leave me a comment.
You can also reach me at:
Twitter: @alc2022
LinkedIn: antoniolinaresc
Great write up.
I do agree with a lot of what has been said here, honestly my main issue with the metaverse thing has been the poor communication about what they seek to do with it.
I remember a point blank question on the earnings call about the metaverse and they basically said they didn't know what it would bring in 2025.
Honest? Maybe, but any owner of Meta would be concerned about a company pouring billions down a path with such uncertainty.
That said, the price has taken out a lot of the risks.
"the things that have done the most good for humanity have also done the most harm"
Reminds me a lot of what Ben Thompson frequently says, that technology is neither intrinsically good nor bad, just neutral, but people will take it in both directions. Money, TV, internet, gaming, social media, crypto
Fantastic piece as usual