Meta Platforms’ “year of efficiency” is almost over. The social media group’s cost-cutting drive has helped to lift the stock out of its slump in a revival that has more than doubled the company’s market value since January. Its next step is trying to prove that it can capitalise on the hype around artificial intelligence and fuel a longer-lasting rally.
In practice, that means the return of cost growth. Total operating expenses are set to rise to between $94 billion and $99 billion next year for the owner of Facebook, Instagram and WhatsApp, edging up from the $87 billion to $89 billion the company anticipates this year. Spending will be focused on the hardware and data centres needed to expand its AI technology, as well as a shift in hiring in “higher-cost technical roles”. That will come at the expense of other projects that sit outside the technology of the moment.
Yet to translate AI hype to a sustainably higher rating, Meta will need to show tangible proof that its social media platforms can actually monetise machine learning.
Meta still trades at the lowest valuation of any of the so-called Magnificent Seven technology groups that have powered the American stock market this year. Its shares are priced at just under 18 times forward earnings, a market discount to the likes of Apple, Amazon and Nvidia.
That is easier to justify in some cases than others. Take Nvidia, which has announced blowout sales growth by making the chips needed to power generative AI. Or even Amazon, which investors hope can capitalise on the AI boom through its cloud computing platform. But for Apple, where the slowdown in consumer spending in big markets such as the United States and China has brought sales growth to a crawl pace, it is harder to account for.
Advertisement
The relative quality of Meta’s income stream is one obstacle for gaining a higher rating. There is little diversification. Advertising generates almost all its revenue, which means it is dependent on the spending power of businesses and, in turn, consumers.
Shaky recovery
Third-quarter figures showed some signs of recovery. Revenue was up 23 per cent, ahead of market expectations, in an echo of better ad spending reported by Snap and YouTube. Spending per advert was down 6 per cent, but that was less than half the 16 per cent decline in the second quarter.
Reels, its short-form video format, is no longer cannibalising revenue. The video function, which is harder to monetise than Facebook and Instagram feeds, is expected to start adding to revenue next year.
Efforts to slim down the structure of the company, which left headcount down by about a quarter year over year during the September quarter, also have boosted profit margins, but the outbreak of fresh global political unrest could put a halt to the improvement in sales. The onset of the conflict between Israel and Hamas coincided with a pause in advertising spending by some of its clients. Revenue guidance for the fourth quarter has been set at a range of between $36.5 billion and $40 billion, wider than normal, to reflect that uncertainty. At the midpoint, that outlook would be 19 per cent higher than last year, but was still below the consensus forecast.
Next year, the sales comparatives will be less flattering than they have been over the present 12 months. In the third quarter, a rebound in spending from online retailers was the largest contributor to year-on-year growth, while the gaming sector was also called out by management. Both were driven by higher demand from Chinese companies trying to appeal to consumers in other markets. A further slowdown in the powerhouse economy could sap spending power for the companies based there.
AI bets
Advertisement
Meta is now trying to convince the market that it is not being left behind in the race to monetise AI. It had an early start, setting up its Fundamental AI Research lab in 2013, but more recently it has been overshadowed by Open AI’s ChatGPT, backed by Google’s parent Alphabet, which thrust generative AI into the public consciousness.
Amazon and Microsoft also have established a clearer narrative around the potential for their cloud computing businesses, which dominate the market, to provide the space for developers to build and run their “large language” models. Both have taken stakes in AI companies that are in exchange demonstrating proof of concept by using the technology group’s cloud platforms.
Meta has developed its own large language model, LLaMA, which is available for free and commercial use. The idea behind making models open source is to encourage take-up and the amount of information put into them, which should give them more powerful capabilities eventually.
Thus far, AI has centred on the recommendations seen by users on Facebook and Instagram, which crucially are designed to improve targeted advertising. Meta’s use of generative AI is still in its infancy. Meta AI was rolled out only weeks ago, a chatbot assistant that provides individuals and businesses with real-time information and can generate photorealistic images.
Investing in integrating AI into its core social media platforms is easier to substantiate than its Reality Labs business. The group will continue to pour cash into the so-called metaverse, an immersive world accessed by virtual-reality goggles.
Advertisement
The pet project of Mark Zuckerberg, Meta’s boss and founder, continues to be a money pit. It announced another $3.74 billion in operating losses in the third quarter and will push further into the red for the year overall. Losses are expected to step up again next year and there has been no guidance on when those might peak or when it will break even.
Meta will not be drawn on its fourth-quarter expectations for unit sales or revenue for the Reality Labs business. Whether its smart glasses and goggles make the jump into mass adoption, enough to justify the huge outlay, is highly questionable.
Investors would be better served by Meta cutting its metaverse bet and diverting cash towards strengthening the competitive position of its platforms.
ADVICE Hold
WHY The shares are cheap against technology peers, but sales growth could get harder and it is a less clear AI play
