Can Meta keep growing?

The second phase of the business’s life began in October 2021, when Zuckerberg announced that he would be rebranding Facebook as Meta. At the time, the pandemic was forcing everyone to work, shop and socialise at home. If ever there was a perfect environment created for an online social networking business to succeed, it was this one. At $323 (£252) a share, the company’s market value was up 70 per cent over the past two years. It seemed as though there was little that could go wrong, but the fall wasn’t far away.

It was the publication of annual results for 2021, released in March last year, that confirmed investors had lost faith in Meta’s growth story. Issues had been bubbling for some time: user growth was slowing, the virtual reality division was losing an increasing amount of money and Apple’s (US:AAPL) incoming privacy rules were about to upend Facebook’s advertising model. Fourth-quarter figures brought things to a head. Reality Labs, the virtual reality division that forms the centre of the company’s metaverse plans, was still losing billions of dollars, while for the first time ever Facebook users declined month on month.

Meanwhile, analyst attention on the company earnings call focused on Apple’s decision to allow users to prevent data tracking on iPhones. Meta estimated that without being able to target ads based on people’s browser history it would lose $10bn (£7.8bn) in revenue the following year. 

The reasons why Facebook’s daily active user growth had faltered were less concrete. The company speculated that it was a combination of Covid-19 having pulled forward user growth into earlier periods, an increase in data pricing in India, and greater competition from TikTok; the Chinese-owned social media company was mentioned six times on the results call. While total user numbers only fell 0.5 per cent, the drop was seen as symbolic of the fact that Meta was no longer a growth stock. At one point in 2022, the company's 2024 price/earnings ratio (PE) fell below 10. For the first time in its history, Meta was cheaper than the S&P 500.  

Meta strikes back

In a way, the negative outlook was proved correct, albeit not necessarily for the reasons first thought. This February, it announced that 2022 revenue dropped 1 per cent and that rising costs had pushed down earnings by 38 per cent. New chief financial officer Susan Li blamed weak fourth-quarter (Q4) advertising demand on “the uncertain and volatile macroeconomic landscape”.

Rising energy costs, borrowing costs and general consumer weakness meant companies had been pulling back on their advertising budgets. Alphabet (US:GOOGL) also saw its advertising revenue drop 4 per cent that year. There is an argument that these tech companies have become so big, and rely so much on advertising, that they have become proxies for the global economy.

However, there were reasons for optimism in the full-year results. The company took $4.2bn in one-off restructuring costs, which decreased earnings in the short term but promised longer-term profitability. This included paying severance costs for 11,000 employees and changing data centre architecture so they could switch between AI and non-AI workloads.

Zuckerberg refers to 2023 as the “year of efficiency”. In February, Meta lowered its 2023 expense forecast by 5 per cent, from $100bn to $95bn. After ploughing money into virtual reality, signs that the chief executive could be more responsible with his cash have cheered investors this year.

Zuckerberg is the last remaining founder-chief executive of a large technology company, and with special voting shares he can make decisions with little oversight. When the economic situation started to turn, this was used as a stick to beat him with. However, his ability to make quick decisions is now being seen as a strength. “Alphabet and Meta are equally bloated companies, but Zuckerberg has been much quicker to make the tough decision to let people go,” says Denny Fish, portfolio manager responsible for Janus Henderson’s global technology strategy.

In three years to 2022, Microsoft (US:MSFT), Alphabet and Meta all increased their headcount significantly. In 2017, Meta’s headcount was 25,105 and by 2022 it had risen more than threefold to 86,482. In that same period, its operating margin fell from 50 per cent to 29 per cent. Tech companies are supposed to become more profitable with scale, but Meta was hiring so quickly the opposite was happening.

In the past year, Zuckerberg has moved quickly to correct this mistake. The 21,000 people Meta has laid off since November amount to 25 per cent of the total workforce; by far the most aggressive redundancy programme among the big tech businesses. Alphabet cut 12,000 people, but due to its size that was just 6 per cent of the total. Meanwhile, Microsoft laid off 5 per cent and Amazon 2 per cent.

Analysts and the stock market have responded positively. Analyst consensus, according to FactSet, is for the operating margin to increase to 30 per cent in 2023. While still a far cry from the 40 per cent average seen over the five years to 2021, a rise this year would be only the second increase in the past five. Meanwhile, the 2024 earnings per share (EPS) forecast has risen from $10.3 at the end of last year to $14.7 at the time of writing. “Meta was one of the most profligate companies, but then it got a new chief financial officer and completely changed its image to the market,” says Polar Capital fund manager Ali Unwin. Li joined Facebook in 2008 and has worked her way up through the business over the past decade. 

AI and the advertising recovery

Cost-cutting is one thing, but to justify its multiples Meta needs to grow healthily, too. To do this, it is relying in part on AI-driven enhancements to its core advertising business. 

The investments needed to drive both the metaverse and AI advancements boil down to the same thing: server and computing power. Meta has long been investing heavily on this front. In the past two years, its capex totalled a cumulative $64bn, up from $34bn in the previous two years. Li has been trying to bring down operational costs, but she has no interest in lowering capex, and the company has guided for this spend to be between $30bn and $33bn this year. “This outlook reflects our ongoing build-out of AI capacity to support ads, [Instagram and Facebook services] Feed and Reels along with an increased investment capacity for our generative AI initiatives,” the chief financial officer said on the last results call.

This investment was made to counter the damage caused by Apple’s privacy shift. Previously, Meta relied heavily on tracking users across the web to know what to advertise to them on Instagram and Facebook. For example, if someone looked at a shoe on the Nike website, it would appear again on their Instagram feed. That ability was lost with the introduction of Apple’s changes.

Meta’s response was to buy lots of graphics processing units (GPUs) to increase its AI capabilities. To improve its advertising efficiency without the use of website tracking, Meta turned to its own proprietary data; with billions of users interacting with a range of different content every day, it has a lot of information to use. Now, the company effectively buckets users based on what they do on Instagram and Facebook, then sends them ads it thinks they are likely to interact with. “Really most of our AI investment today is supporting the building of [this] Discovery Engine and ranking unconnected organic content,” said Li.

This sounds simple, but it takes a huge amount of computing power. So far, the investment has paid off. In the first quarter of this year, advertising revenue rose 4 per cent to $28bn, the first year-on-year quarterly increase since the end of 2021. This improvement seems modest, but it must be seen in the context of a wider digital advertising market that has been weakening over the plast 18 months. Snap (US:SNAP), another social media company which makes its revenue from advertising, saw its sales drop 7 per cent in the same period.  

Ironically, there is a belief that Apple’s privacy changes might ultimately benefit the large internet companies such as Meta, Google and Amazon. Previously, tracking meant relevant ads could be put on any website regardless of their size, and therefore even small websites could sell high-return advertising space. The argument is that it’s now necessary to have a huge amount of proprietary data and AI capabilities to serve relevant advertising.

Amazon and Google are still best placed on this front, because people search for exactly what they want to buy on those websites. Meta is not an ecommerce service, so its data is not quite as valuable. However, with its new AI capabilities, it had been able to do a lot more with what it has. 

Two ways it can do this are via the Advantage+ advertising tool and the Reels social media feature. Advantage+ uses AI to serve custom-made adverts to targeted audiences at a lower cost. The assumption was that Meta's ballooning R&D costs were going towards virtual reality, but a lot of it was being channelled into these two products to counter the challenges presented by TikTok and Apple.

If Advantage+ can bring down the cost of advertising by removing marketing agency fees, then it could create a lot more customers for Meta. “The creation of content and copy is often the biggest barrier for small companies to advertise,” says Unwin.

Reels is a short-form video product launched on Instagram to repel the threat from TikTok. Last year, the concern was that Reels’ growth was damaging engagement with the parts of Instagram that are more lucrative for Meta – Feed and Stories. But the company now says that Reels will be revenue-neutral by the end of 2023, and its increased user growth suggests it has been successful in stemming the TikTok threat – as well as driving engagement with Instagram as a whole.

Advertising growth is also being helped by new ad formats, such as click-to-message (CTM), which allows users to contact companies directly via Meta services such as WhatsApp and Facebook Messenger. Analysts at RBC estimate that CTM – over half of whose advertisers only use this format – will contribute 250 basis points of Meta’s 8-10 per cent forecast annual revenue growth in 2023 and 2024.

The metaverse

Alongside AI and advertising, the other big area of investment for Meta has been virtual reality. This part of the business – known as Reality Labs – has been a huge lossmaker over the past five years, adding to the scepticism over the concept itself. It was in 2014 that Meta launched its virtual reality ambitions with the acquisition of Oculus for $2bn. Since then, it has poured money into the product and completely rebranded the company, all without seeing much in the way of sales growth. Last year, Reality Labs’ revenue fell 5 per cent to $2.2bn, while its losses widened 35 per cent to $13.7bn.

Investors have been sceptical about Reality Labs because they don’t see an obvious use case for the technology. However, there is some optimism that Apple’s entrance into the market might help here.

At the start of June, Apple announced its Vision Pro headset. A few days earlier, Meta announced the launch of the Meta Quest 3 headset. Zuckerberg is nonetheless reluctant to be drawn into head-to-head comparisons. The Meta Quest is priced at $500 whereas Apple’s Vision Pro will cost $3,500, and Zuckerberg seems to think his product, and the metaverse it feeds into, will prove better for socialising.

So far, Zuckerberg has been good at keeping up with the future. But whether it be the launch of the iPhone (which threatened Facebook's desktop computing-focused business model), privacy changes or now the Vision Pro launch, the challenges to his company often come from Apple.

Whether Zuckerberg is capable of competing with the world’s most successful company directly, in the world of virtual reality, is uncertain. But if either of these products does eventually take off, they will do so over a period of years rather than months. The improvements elsewhere in Meta’s business have at least bought it some time on this front. In the past two years, Reality Labs has made an operating loss of $24bn without producing any revenue growth. Thanks to its massive advertising business, this is manageable. Last year, it still made $50bn in operating cash flow and, with its new AI capabilities, analyst consensus expects this to rise to $80bn by 2026.

Valuation

The company’s ability to produce excess cash flows has been at the forefront of its success over the past decade. Free cash flow per share growth over the five years to 2019 (ie pre-pandemic) stood at 40 per cent a year. Over the coming five years, analyst consensus puts that growth rate at 10 per cent – a big decline, but still a healthy level of expansion considering free cash flow in 2022 stood at almost $20bn. And Meta’s cash generation is particularly valuable at the moment. The costs associated with AI, from both a hardware and development side, are massive – and with US interest rates now above 5 per cent the ability to fund investment internally gives Meta, Alphabet and Microsoft a moat between themselves and potential competitors. “Ultimately, [the likes of] Snap just don’t have the same scale and capabilities as Facebook to compete in AI, this is an era where size matters,” says Fish.

But unless Zuckerberg’s bet on virtual reality proves correct, the days of explosive growth are behind the company. In 2022, Facebook’s daily active users grew just 4 per cent to 2bn. Across all its apps it now has 3.74bn users. At this scale, it’s tricky to keep bringing in new users, and meanwhile digital advertising as a share of the total advertising market has started to flatline. Over the past year Meta has taken market share from the smaller advertisers, but consistently scaling up from current levels will still be a challenge.

Between 2012 and 2021, top-line growth never fell below 22 per cent. In 2013 and 2014, when it introduced its new mobile advertising products, growth was above 50 per cent. Apple had put computers into everyone's hands, digital advertising was growing rapidly and Facebook was one of the big beneficiaries. Its valuation reflected this. In 2014, Facebook was trading on a forward PE ratio of over 50, compared with just 15 for the wider S&P. 

Things have changed since. US economic growth has slowed to a halt and so has Meta’s: see last year’s 1 per cent revenue drop. But next year it is forecast to grow around 8 per cent, and thereafter hover around the 10 per cent mark. For the past few years, Meta’s valuation has more or less tracked the wider S&P 500, and while these revenue figures don’t justify a gigantic multiple, they are well above those forecast for the index as a whole, as are earnings per share forecasts. With those expectations baked in, the key will be to grow further. For Denny Fish at Janus Henderson, the company now looks like good value. “In some twisted way, Apple’s rule change actually benefits the mega-cap stocks such as Meta and Google, and for the first time these companies are trading on reasonable valuations,” he says.

The company is unlikely to post extravagant figures for a long time, but the worst-case scenario has gone: it has bolstered its competitive moat and is still capable of churning out growth. Apple’s privacy rules did not kill the business, the threat of increased regulation has not hurt it so far, and TikTok’s threat is receding – pressure from the US government is also helping on this front. Meta might not be a high-growth stock any more, but in the past year it has seen off another set of existential risks. Resilience is a strength few businesses have and, in a rapidly changing world, it is going to be more important than ever.