Goldman’s 50-Stock Portfolio Of Under-The-Radar Long-Term AI Winners

Artificial intelligence has already launched a few big-name tech stocks into intense limelight. But because every investor and their chatbot of choice knows about them, you’d need to lay down stacks of cash to get involved. So Goldman Sachs wanted to find more firms that could really plump up profit using AI in the future – crucially, ones that investors aren’t all het up about yet.

What did Goldman factor in?

Goldman focused on AI’s impact on long-term headcount – that is, the number of human working hours that could be wiped out, à la “Charlie And The Chocolate Factory”. Check out the chart below: while only 22% of Fortune 500 CEOs surveyed think the tech will reduce working hours in the next year, 74% expect it to within the next five.

The percentage of Fortune 500 CEOs that expect AI to impact their workforce within the next year, and within the next five years. Source: Goldman Sachs.

Bear in mind, AI won’t affect all firms equally. Goldman analyzed nearly 1,000 US job roles, segmenting them by the difficulty and importance of their respective tasks. The more simple and less important tasks are the ones Goldman expects AI to take over first.

After that, the big bank filtered out firms that aren’t heavily impacted by labor costs. After all, AI could displace all of a firm’s workers, but if its existing wage bill isn’t that big, it won’t matter all that much to profit – or investors.

How did Goldman work out the change in potential profit?

That’s all interesting stuff, but investors want numbers. So to quantify these expectations, Goldman priced up two scenarios. The first: companies use AI to replace all the tasks that make sense, replacing humans with machines and dramatically shrinking wage bills. The second: companies equip their workers with AI tools, roughly maintaining wage bills but increasing efficiency and, in turn, revenue.

Goldman then took the average of the two routes: the shaved costs of firing workers or the higher revenue from AI-induced efficiency. Then, the big bank ranked firms by the change in profitability under this scenario.

Stats calculated, Goldman whipped up a 50-stock portfolio, equally split so each one made up 2% of the lot. Note that the portfolio’s been arranged so the proportion of each sector is aligned with the Russell 1000 index. That means if you put 2% of your cash into each of the stocks on Goldman’s list, you’d match the industry weightings of the index. See, if – without this specification – you just picked up the top 50 high-scorers, you’d end up with a heavy weighting in a couple of industries that could be more easily automated, like tech. So by building a portfolio that replicates the index’s sector weighting, you’re more likely to get a diversified bunch of investments – a less risky spread, at least in theory.

Which stocks made the list?

Check out the table below: that’s Goldman’s final roster, with each firm’s estimated profit improvement.

Goldman Sachs’s 50-stock, equally weighted portfolio of long-term AI beneficiaries, their sectors, and the potential upside to profit from AI adoption. Source: Goldman Sachs.

Let’s pick out the median value of the bunch as an example – the middle, not the average, to eliminate outsized influence on either end. Goldman reckons 36% of this stock’s labor force would be impacted by AI, and existing labor costs would be worth 36% of the firm’s sales. (Compare that to the Russell 1000’s median stock: 33% of its labor force could be disrupted, but labor costs only make up the equivalent of 14% of sales.) When Goldman factors in the potential profit uplifts we mentioned above, either a reduction in wage bills or an increase in efficiency and revenue, the list’s median stock would emerge with a potential 72% profit plump after an artificial helping hand.

Recall that CEO survey from earlier on. That’s leaders of major companies we’re talking about, and they’re clearly not expecting AI to overhaul their companies in the immediate future. So you wouldn’t expect investors to be pricing major tech changes into the stocks of those firms either – not yet, anyway. This, then, could be your chance to get ahead of the game before prices pick up. Case in point: Goldman points out that an equally weighted portfolio of today’s most prominent AI pioneers, the likes of Nvidia and Microsoft, has outperformed the equally weighted S&P 500 by 62 percentage points. Goldman’s shortlist of long-term hopefuls, meanwhile, is only six percentage points above the index. For me, that could signpost a chance to get a foot in the future’s door for less.