AI, Big Tech, And All The Big Questions

ChatGPT is everywhere right now. It’s no surprise then that it’s had a lot of people wondering whether artificial intelligence (AI) is the next massive tech wave, and, if so, which companies are best positioned to ride it. I hosted a busy Ask An Analyst session last week, about this and other Big Tech topics. Here are some of the questions you had, along with my answers, edited for length…

Q: We might argue that "AI is the future" with a high probability. My question is, do you think that the current tech leaders (such as MSFT and META) will still have the edge in the next five to ten years, or is there still a chance for new AI startups to find their way to the top – and thus represent good investment opportunities right now?

A: Look, despite all the hype, AI’s still very early in its development, and investors have time on their side. That might seem an odd thing to say given what looks like an AI arms race. But here’s how things tend to go with these tech waves: after the initial hype, the buzz tends to die down, before eventually ramping again.

I expect Big Tech to lead AI development, but there are bound to be important AI companies that haven’t even been started yet, and companies that will see AI completely change the foundations of their business in an unforeseen way. I’d be focusing on sectors like healthcare, industrial, or consumer goods – ones that stand to benefit from the technology through efficiency gains or new revenue opportunities. And I’d try to avoid rushing into anything, especially if you don’t like Big Tech’s stocks. It’ll take a long time for mainstream firms to come to grips with AI and figure out where it fits in their business and in our lives. After all, cloud computing was being hyped in the mid-2000s, but you didn’t need to invest in Microsoft until around 2016 to take full advantage: that’s when its shares really started to outperform the S&P 500.

Microsoft’s performance relative to the S&P 500. An upward move in the line indicates outperformance. Source: Koyfin.

Q: How often is AI used by institutional investors to analyze stocks and make predictions? What do you think about AI for stock picking, and what tools would you suggest using?

A: For sure, some bigger hedge funds and asset management firms will be experimenting with AI. I have no idea how it will impact the asset management industry, but it’s bound to affect it in a big way. I’d draw parallels with quant-driven strategies that are now responsible for a big proportion of institutional assets under management. That said, I don’t think AI will replace humans when it comes to decision-making anytime soon. (Said with my fingers firmly crossed). Check out Reda’s recent Insight on AI and investing: he used ChatGPT to analyze Tesla. My takeaway was that the outcome was decent, but only because Reda knew what to ask. So, AI might be able to do some of your research navigation at some point, but you’ll still need a driver at the wheel.

Q: Which tech company today, in your opinion, looks like Meta, Google, and Apple ten years ago? Is AI the big new technology? And, if so, who's most advanced? If not AI, what’s the next big technological advancement we should be looking for? e.g. driverless trucks, trains, etc. Staffless service jobs in retail food outlets? Thanks.

A: I think you’re right to broaden your vision and think about how tech trends like AI will impact other industries. In technology, I’m following the semiconductor industry closely, because all the technologies you mention, including AI, will be very semiconductor intensive. I’m staying aware of the risks, though, as they’re cyclical commodity companies at the end of the day. But when I take a step back and look ahead through the cycles, I think there’s a very good chance that the overall semiconductor industry will be a lot bigger in ten years' time than it is today.

iShares Semiconductor ETF price (blue line) and volume (red and blue bars). Source: Koyfin.

Q: Many of the Big Tech firms have super ambitious ESG targets, but how many of them are forecasted to be on track with those? What are the biggest ESG risks you see for Big Tech companies in the next two or three years?

A: Big Tech firms tend to score well on environmental, social and governance (ESG) criteria because of their relatively low carbon footprints – they’re services businesses that mostly don’t actually make anything, after all. The biggest risk I see is on the social and governance sides, where Big Tech has a patchy record. Now, the big question for these giants is whether they can be part of the climate solution and whether governments will incentivize them to drive technology forward. Who knows – maybe pressure will come from lawmakers, in a sort of quid pro quo arrangement – something like “we won't seek to curb your monopoly-like power (including AI) if you use your advantage to dramatically impact the climate effort”.

Q: Do you think we’ll see an M&A buying spree in AI, with big acquirers taking advantage of lower valuations? Or will higher interest rates discourage acquisitions?

A: When it comes to mergers and acquisitions (M&A), I think Big Tech is going to be active – as Microsoft has already shown with its multiyear, multibillion-dollar investment in ChatGPT maker OpenAI. These companies (and their shareholders) have become used to rapid growth, and I suspect these firms will want to prop up those growth rates with M&A. From a valuation perspective, small-cap stocks sold off heavily last year and still look attractively valued, relative to their history. Big Tech would likely run into competition issues with heftier deals too. That said, watch out for some pretty chunky adjacent deals, like the Microsoft-Activision one, for example. These behemoths aren’t short of resources: they have a combined $400 billion war chest.

Big Tech has almost $400 billion on hand for acquisitions. Sources: Koyfin, company accounts.

Q: How much will buybacks affect markets in 2023, and how much of it should we expect from the (usual) big players?

A: I’ll broaden this question to net buybacks – or the difference between new stock issued via stock-based compensation (SBC) and the share count reduction from buybacks. Big Tech buybacks have mostly only offset the dilution from SBC over recent years. But with hiring in the industry slowing, there should be downward pressure on SBC. And given the cash resources (and cash production) at Big Tech, I would expect buybacks to ramp up in order to support earnings-per-share (EPS) growth in a lower-revenue environment.

So the combination of a faster buyback pace and lower SBC should mean that the impact is more meaningful in terms of EPS growth this year. Over a long time frame, the net impact (that’s after SBC) of buybacks for the S&P 500 has been less than you’d think: boosting EPS by no more than one percentage point. (Stéphane wrote more about this here.) Going forward, though, I’d expect this to improve. My guess is that net buybacks could juice EPS growth for Big Tech by something like five percentage points over the next couple of years.

Q: Do you think the layoffs across Big Tech are “one and done” (with staffing returning to last year’s levels)? Or do you expect that we’ll see more in the coming year(s) as they reduce costs?

A: I don’t think they’re done yet. The market’s reacted well to job cut announcements so far, and that’s surprised me a bit, I have to say. I’d expected investors to view those cost cuts as an admission of Big Tech running out of growth runway, and to pressure valuations lower. Instead, the market’s applauding layoffs as a display of long-overdue cost discipline.

The prevailing view is that the current downturn in sales growth is only cyclical in nature, and it’ll reaccelerate later in the year. If it does, profit margins on the other side should be fatter after efficiency-driven cost cuts – and that will lead to a big reacceleration in EPS growth. But if slower sales growth is a more permanent thing, it might be a different story. This is one of the reasons why I don’t think Big Tech valuations are going to expand back to the late 2021 levels in the near term, even with all the AI hype.