Meta needs no introduction.
1. Demand
Beat revenue estimates by 1.0% & beat its revenue guidance by 2.0%.
Beat daily active user (DAU) estimates by 5.0%.
As of this quarter, Meta no longer discloses Facebook-only DAUs – only user metrics for its Family of Apps (FOA) overall.
2. Profitability & Margins
Beat EBIT estimates by 3.1%.
Beat $4.33 GAAP EPS estimates by $0.38.
Beat free cash flow (FCF) estimates by 10.1%.
3. Fortress Balance Sheet
$58 billion in cash & equivalents.
$18.4 billion in total debt.
Share count fell by 1.7% Y/Y.
Headcount fell 10% Y/Y, but continued to rise Q/Q.
4. Guidance & Valuation
For the 2nd quarter, the midpoint of Meta’s revenue guidance missed by 1.2%.
For the year, it raised its operating expense (OpEx) guide by 1.0% & raised its capital expenditure (CapEx) guide by 11%. It also told us that 2025 CapEx would rise Y/Y vs. 2024.
Meta trades for about 20x earnings with earnings expected to grow by 35% Y/Y. With the 1% OpEx boost for 2024 and the Q1 EPS beat, I don’t anticipate earnings estimates falling much after this report. Conversely, it trades for about 21x 2024 FCF with FCF expected to grow by 13% Y/Y. I would anticipate FCF estimates being revised lower with the large CapEx raise.
Love this visual?
My good friend Carbon Finance sends out a weekly visual newsletter with the most important infographics, insights, and insider trades.
It’s completely free and only takes 5 minutes to read.
Join 10,000+ investors and subscribe here now.
5. Call & Release
The Next Investment Cycle:
Internal optimism surrounding Meta’s ability to create world-class AI models and services is emboldening leadership to lean in. It’s planning to invest even more aggressively in building out AI infrastructure to ensure no capacity bottlenecks. That can be seen in the CapEx guidance raise. Its view of its role in the GenAI wave is becoming more and more positive. STILL, the coinciding boost to investments is coming before these newer products are ready to be real revenue drivers. That timing mismatch will be a margin drag in 2024 and perhaps 2025 too.
While investment levels will ramp, they’ll ramp with Meta’s newfound commitment to efficiency and investment scrutiny. It will continue to fixate on model training/inference efficiency, vertical integration of some chip use cases and “carefully managing headcount as well as overall costs. This will not turn into a repeat of its 2021-2022 mistakes. The focus on margin preservation is NOT being thrown in the garbage like it was then.
To me, this is the source of the after-hours weakness for Meta… but it’s a source of weakness that I’m actually quite excited about. I haven’t been able to add to my Meta stake in over a year, and will be tempted to do so after this report and dip. We have countless examples of investment cycles being temporary and eventually leading to explosive incremental revenue and profit. It fixates on creating incremental value… driving massive scale and THEN monetizing. It never immediately monetizes.
It’s the exact same playbook that it always runs… from the Metaverse… to Reels… to WhatsApp… to Stories and to the original Facebook product. While this current investment cycle will likely be more intense than others, there are zero teams and firms on the planet that I am more comfortable betting on than this one. It has half of the planet on its apps; it has a fortress balance sheet; it is a free cash flow monster; it is already seeing uniformly positive signs that these investments will bear significant fruit. Zuck has delivered time and time again... and I expect this to be no different. I will personally trust the process.
For now, Meta AI (powered by its newest Llama 3 foundational model) recently was broadly released across Meta’s apps. Scaling this in 2024 is a key priority as it pushes to make this more naturally conversational and utility-building. It sees this becoming the most broadly used and best free assistant in the market.
“Success is leading me to believe we need to invest more in advanced models and services… Optimism has grown quite a bit… Initial signs are positive… We will need to grow investments meaningfully before we generate much revenue from these newer products… We’ve proven we can build leading models, so we’re going to go for it.”” — Zuck
Where AI is Already Helping Financials:
AI is already helping Meta’s apps become more engaging. So while Meta AI and other new bets may take longer to enjoy material revenue, its most mature work here (or “core AI” work) is already helping.
For example, 30% of Facebook content viewed is via these AI rec engines vs. 15% a few years ago. AI-based discovery content also now makes up more than 50% of all Instagram engagement for the first time ever. Better AI models mean better targeting and more engaged users. It sees more opportunity to use AI to sharpen the Reels business while consolidating recommendation engines to cut costs and boost efficacy. For example, combining the Reels rec engine and live video rec engine on Facebook is already driving 8% to 10% engagement gains.
AI is also helping to automate campaign creation and return on ad spend (ROAS) for buyers. It’s helping to ease data onboarding and campaign creation within Advantage +. For example, model advancements here unlocked better segmentation and audience suggestions to drive a 28% reduction in cost per click. Advantage + Shopping and App campaigns enjoyed 100% Y/Y growth, and this was likely a material contributor.
GenAI is also already being directly infused into its Meta Lattice product to help prioritize and select placements. The new architecture is pulling from larger, aggregated models rather than smaller models to cut costs and boost efficacy — just like with the previous Facebook example.
Within Reels specifically, AI is a material piece of it continuing to improve performance and ad load like it expects to. It has helped Meta shift ad loads and formats to more lucrative and less disruptive avenues within Reels.
Simply put, AI is a vital instrument in Meta’s ability to maximize ad performance. AI model training and inference gains will continue to be leaned on to power constant model improvements over time. These improvements mean more time spent on its apps… and that means more ad revenue.
WhatsApp and Threads:
WhatsApp user and engagement growth in the USA continues to be a standout for Meta. Great news that many previously thought was infeasible. Business messaging on that app is helping drive 82% Y/Y growth in its other revenue segment to reach $334 million this quarter.
Threads now has 150 million monthly active users (MAUs) vs. 130 million Q/Q.
Costs:
Gross margin expanded significantly from 79% to 82% Y/Y. This was mainly due to lapping inventory valuation adjustments within Quest. All OpEx line items fell as a percent of revenue.
Notably, a 13% effective income tax rate vs. 22% Y/Y helped EPS growth. With a 22% tax rate this quarter, it would have earned $4.21 in EPS (almost 100% Y/Y growth), but the lower effective tax rate was as expected by analysts. I just provided this data point for added context.
Ad Growth Metrics:
Impressions rose by 20% Y/Y.
Price per impression rose by 6% for what was by far Meta’s best result in a long time. It’s improving across all geographies. This was driven by healthy demand, which is great news for all advertising players.
North American ad revenue rose by 21.6% Y/Y.
EU ad revenue rose by 32.9% Y/Y.
APAC ad revenue rose by 24.6% Y/Y. Chinese seller ad demand remained very strong.
Rest of World ad revenue rose by 39.9% Y/Y.
Final Notes:
The Ray Ban smart glasses continue to do much better than expected. This is giving the team confidence that this revenue driver will become material before previously expected.
The Meta Horizon operating system is being open sourced (as previously discussed).
When asked about a potential TikTok ban, the team rightfully declined to comment. Good job.
Reality Labs and AI investments are beginning to converge as things like the Ray Ban multi-modal META AI assistant can field questions without a phone.
6. Take
Meta the stock had been on a historic run heading into this report. Sell-side estimates were zooming, buy-side targets were even bolder, expectations were sky-high and the bar for success was wildly elevated. This quarter needed to be absolutely perfect to be handsomely rewarded. It was a good quarter… but the setup meant good simply was not good enough. That matters for people with options expiring this Friday. It does not matter to a boring long term investor like myself.
The degree of the CapEx raise was a surprise, but one that I’m selfishly ok with. And considering it’s coming as a response to better than expected traction and opportunity, I’m even more ok with it. Risks surrounding Meta’s business have faded or vanished in the last three years. It overcame Apple’s data sharing restrictions. It overcame its most intimidating competitive threat to date without any bans. It showed that it can flex margins while investing in longer-term projects. It proved that user growth was far from over. So what was left to worry about? Nothing. And when that happens, even the slightest inkling of a new risk will be routinely amplified and punished. That is what I see happening here.
As I said above, I haven’t been able to touch what is my highest conviction holding in over a year. This new risk is exactly what I needed to get that opportunity. The thesis is firmly intact; I’m more confident in Meta’s ambitious investment periods yielding elite returns than any other firm in existence. Want to spook investors with talks of a multi-year CapEx cycle? Cool. Those thinking in years and not minutes have seen this many times before. Those willing to bet on this team have been uniformly rewarded. This will be more of the same in my view… and I’m gearing up to add to what is already my largest holding.
Match Group may be on the chopping block before its quarterly report to create the liquidity to do just that.
I remember people being skeptical when Meta was hitting its lows at $88 less than 2y ago... back then it was also due to increased CAPEX along with the spend in VR. The CAPEX from that time was supposed to improve Instagram Reels, but because they bought twice the capacity needed... they were able to capitalize on the opportunities with gen-AI and ended up training a state-of-the art LLM (Llama, now on its third version). Source: https://www.dwarkeshpatel.com/p/mark-zuckerberg
The thing is, they have the best people in the world working on and researching AI. Their CAPEX in the past and investment decisions have in general worked out. But still "investors" in Meta seem so fickle as to sell off close to 20% post earnings.
Oh well, their fear & short-termism is my opportunity. I picked up another 100 shares at an average price of $409 in the after hour.