Following our April coverage applauding Meta Platforms’ (NASDAQ:META) strong Q1 performance — a signal of a potential business turnaround —, we’ve seen the company’s shares continue their impressive trajectory. This uptick aligns with Meta’s enhancing business parameters and, notably, its remarkable advances in the field of artificial intelligence (AI). With this article, we seek to delve deeper into Meta’s AI capabilities, exploring how they bolster its competitive edge and fuel growth. Concurrently, we will carry out a comprehensive financial and valuation examination.
From its initiation in 2013, Meta’s journey into the AI landscape has been an intriguing blend of innovative strides and strategic maneuvers. This journey has led to significant contributions to the industry, including the creation of PyTorch, a premier AI programming language. Meta’s unique strategic vision, reflected in its dedication to open-source AI models, is instrumental in fostering community growth and setting industry benchmarks. When combined with its emphasis on AI-propelled user experience and revenue growth, Meta’s path paints a compelling picture of its future aspirations.
Meta’s AI Journey: An Introduction
Meta commenced its journey into the artificial intelligence (AI) landscape in 2013 with the induction of Yann LeCun as the head of their AI research lab. Since then, it has made remarkable strides. One such significant contribution is the development of PyTorch, a leading AI programming language.
PyTorch’s impact on the AI industry has been profound. It is now widely adopted by AI researchers and developers worldwide due to its flexible and intuitive design, which accelerates the AI model development process. This has been pivotal in fostering innovation and advancements in AI.
In a remarkable display of community support, Meta has also publicly shared over 30 AI models and frameworks. These include recent developments such as Large Language Models (LLaMA) and computer vision models. This open-source strategy democratizes access to advanced AI tools, catalyzing progress across the broader AI community.
AI has become a transformative force within Meta, reshaping the user experience from a social graph-oriented approach to one that is predominantly driven by an intelligent recommendation engine. Moreover, we acknowledge the potential of Meta’s burgeoning suite of AI-powered advertising tools. These tools are expected to counterbalance the effects of Apple’s App Tracking Transparency (ATT) policy and fuel a resurgence in revenue growth.
Yann LeCun, at the helm of the Fundamental AI Research Lab, focuses more on the academic aspects of Meta’s AI initiatives. Meanwhile, on the product front, we see Ahmad Al-Dahle leading the charge. Al-Dahle, an Apple veteran, reports to Head of Product Chris Cox, who returned to the company in 2020 and directly reports to CEO Mark Zuckerberg.
Understanding Meta’s Position On Open-Source AI
In our analysis, Meta’s consistent dedication to the open-source community clearly positions it as a leader in the field. The company’s practice of open-sourcing its AI models not only demonstrates its commitment to transparency but also catalyzes advancement within the broader AI community by allowing researchers and developers to experiment with, refine, and build upon their work.
The recently released LLaMA LLM model and the groundbreaking visual models like Segment Anything, DINOv2, and Animated Drawings are testimonies to Meta’s drive toward facilitating communal growth. Not only do these models push the boundaries of what is possible in AI, but their release to the public incites innovation and could lead to unforeseen breakthroughs, benefitting the entire tech ecosystem.
Meta’s open-source strategy is discernibly different from other tech behemoths like Alphabet (GOOG), Microsoft (MSFT), and Amazon (AMZN). Rather than making its software infrastructure proprietary, Meta has found value in creating industry standards through their open-sourced tools. Essentially, Meta is fostering an ecosystem of shared advancement where everyone benefits — a distinct contrast to the more insular, competitive strategies of other companies.
One of the most notable examples of this strategy is the development and open-sourcing of PyTorch, which has now become an industry-standard tool. By offering this tool to the broader developer community, Meta has built a foundation where any improvements or innovations made by external developers can easily be integrated into their own systems, leading to a symbiotic relationship that enhances the overall efficiency and effectiveness of the AI development ecosystem.
However, it’s important to clarify that not all of Meta’s work is or will be open-sourced. The company is aware of the need to create unique value within its own products. They seem to have struck a good balance, though, between contributing to the open-source community and maintaining proprietary technology that sets their products apart.
In our view, Meta’s commitment to open source is more than just a contribution to the tech community; it’s a calculated move that presents numerous strategic benefits for the company. Meta’s open-source strategy acts as a powerful magnet for top talent. Engineers and researchers prefer to work with tools and platforms they’re familiar with. By open-sourcing their AI models and tools, Meta is essentially providing potential hires with a playground where they can learn and innovate using Meta’s technologies. It’s an ingenious way of nurturing the next generation of AI professionals who are already accustomed to Meta’s ecosystem.
Moreover, it’s clear that Meta’s open-source strategy isn’t just about promoting goodwill or establishing industry standards—it’s a clever approach to staying at the cutting edge of AI innovation. By opening their technologies to the world, Meta effectively leverages the collective intelligence of global researchers and developers. Each innovation made within this community using Meta’s models and tools is a potential addition to the company’s technological arsenal, turning the global open-source community into a de facto research and development wing that operates continuously and is driven by passion rather than profit.
Another significant upside of this approach is the opportunity to achieve scale efficiency. By encouraging the widespread use of their tools, Meta can spread the costs related to development, updates, and troubleshooting across a larger user base. This not only reduces Meta’s own expenses but also drives down costs for all users, thereby enhancing the tools’ accessibility and attractiveness, which in turn fuels the further expansion of the user base.
Q1 Earnings Reveal Treasure Trove of Insights on AI Capabilities
Meta’s Q1 2023 earnings call has unequivocally illustrated the central role artificial intelligence plays in the company’s strategic vision. Notably, two parallel technological waves underpin Meta’s growth roadmap – the present and dominating wave of AI and the anticipated wave of the Metaverse. These insights offer shareholders a clear view of Meta’s strategic path and, in our opinion, provide reassurance that the company is on solid footing.
The company’s vast recommendation and ranking infrastructure, one of the twin pillars of its AI strategy, impact all product lines. The benefits are already becoming visible. In fact, revealed during the call, AI now recommends more than 20% of the content in Facebook and Instagram feeds from non-followed sources, and an impressive 40% across Instagram overall demonstrates an effective integration of AI into Meta’s platforms. However, we feel that Meta should pay close attention to ensuring this AI-powered content remains relevant and unintrusive to avoid a potential backlash from users who value their curated social feeds.
Where Meta’s AI investments truly shine is their impact on user engagement and revenue generation. With the launch of Reels, AI recommendations have driven a significant 24% increase in time spent on Instagram, marking a notable win for Meta. Monetization efficiency is equally impressive, with a quarterly increase of over 30% on Instagram and over 40% on Facebook. A sevenfold rise in daily revenue from Advantage+ shopping campaigns over six months is a clear sign that Meta’s AI investments are not only enhancing the user experience but also driving robust revenue growth. In our view, these are tangible outcomes that should reassure investors of the payoff from Meta’s AI strategy.
Meta’s work on generative foundation models, the second key area of AI focus, holds immense promise. We’re particularly intrigued by the potential of these models to transform various apps and services, potentially opening new avenues of user engagement and revenue growth. However, the implementation of these models across diverse platforms, like WhatsApp and Messenger, will need careful execution to ensure a seamless user experience.
We believe the integration of AI agents into business messaging and customer support is an exciting development that holds great promise. As these tools evolve, we anticipate their extension into Meta’s Metaverse work, enhancing the interactive experience of virtual spaces.
Financial & Valuation
Note: All historical data in this section comes from the company’s 10-K filings, and all consensus numbers come from FactSet.
Meta’s Q1 FY 2023 earnings report brought a breath of fresh air to investors with a 13.9% stock surge following the announcement. The company managed to surpass consensus revenue estimates by 3.5%, reporting a year-on-year growth of 2.6% to $28.6 billion. The fact that Meta beat revenue estimates in such a challenging market environment is commendable and signals the resilience of its business model.
However, Meta’s operating margin, at 25.2%, was down from 30.5% a year prior, signaling increasing operational costs. In our view, this is an area that deserves close scrutiny, as declining margins may erode profitability in the long run. Nevertheless, Meta managed to deliver an EPS of $2.20, representing a 19% year-on-year decline but still beating consensus by 8.9%. This beats expectations, demonstrating an ability to deliver shareholder value despite the margin squeeze.
Considering the financial trends, Meta’s revenue growth, driven by a CAGR of 18.2% over the past three fiscal years, highlights the firm’s ability to expand its top line consistently. While the sell-side consensus forecasts a moderation in this growth to 8.0% this fiscal year and 10.8% next year, this is a realistic expectation considering the company’s size and market saturation. Despite this moderation, we remain positive about Meta’s ability to drive revenue growth consistently.
Worrisome, however, is the company’s EBIT margin, which has witnessed a contraction of 9.1% points over the past three fiscal years. On the bright side, the consensus expects Meta’s EBIT margin to expand by 540 basis points this year and by a further 262 basis points next year. This turnaround, if realized, will be a positive signal for Meta’s profitability.
Interestingly, over the past three years, Meta spent 8.6% of its revenue on share-based compensation. This was more than offset by the company’s robust share repurchase program, which led to an 8.7% decrease in diluted outstanding common shares over the same period. However, the EPS growth rate of 10.1% over the past three fiscal years has lagged behind revenue growth, hinting at potential efficiency issues. Future EPS growth, projected at 37.8% this year and 23.7% next year, will be an important metric to monitor closely.
Meta’s free cash flow (FCF) forecast for this fiscal year of $23,453 million represents an FCF margin of 18.6%, down from 28.8% four fiscal years ago. However, given the company’s historical average FCF margin of 27.5%, we believe Meta’s current FCF margin is an outlier and expect it to revert closer to the mean. We also note that Meta’s capital expenditure, at 21.4% of revenue, indicates a relatively high capital intensity, suggesting major investments in infrastructure and business expansion.
With a strong return on invested capital of 16.0% and net cash of $33,681 million, Meta is well-positioned for future growth investments. We are also encouraged by the company’s impressive stock performance over the past year, which saw a 35% point return above the S&P 500.
Meta’s current valuations are mixed. At an EV/Sales multiple of 4.6, EV/EBIT multiple of 14.1, P/E multiple of 18.1, and FCF multiple of 24.3, Meta is trading at a premium on some metrics and a discount on others when compared to the S&P 500. However, we believe that Meta’s FY2 PEG ratio of 1.4, which is at a discount of 16.1% compared to the S&P 500’s PEG ratio of 1.6, is a key metric here. This ratio suggests that despite its higher relative valuation on some metrics, the company’s future earnings growth is not fully priced into the stock, making it potentially undervalued.
Conclusion
The convergence of Meta’s commitment to open-source AI, its strategic talent acquisition, its harnessing of global innovation, and its economies of scale positions it as a major force within the AI landscape. The positive impacts of AI on user engagement and revenue generation, evidenced by the company’s Q1 2023 earnings call, underline the success of this strategy. While there are areas for careful scrutiny, such as the potential backlash from users over AI-curated content and increasing operational costs, Meta’s ongoing investments into AI infrastructure are clear indicators of its confidence in the strategic vision.
Financially, despite challenges, Meta has demonstrated resilience in its business model. While the contraction in EBIT margins warrants attention, the expected turnaround and the company’s ability to consistently drive revenue growth remain positive signs. As Meta continues to negotiate the balance between proprietary technology and open-source contributions, it will be fascinating to witness the unfolding of its journey within the transformative realm of artificial intelligence.
Read the full article here