Will the future of AI boost profits — or just increase competition?
‘‘ Investors have suddenly started worrying that for many companies, artificial intelligence may not prove such a boon as widely assumed. Stock markets have been hit by waves of panic as investors have fretted whether this or that sector could be hurt by the AI revolution.
Software suppliers, data companies, wealth managers and insurance brokers have all come under fire.
Most have recovered a bit and stock market experts tend to agree that the selling was generally overdone. Some will doubtless suffer from AI-driven competition but the consensus is that it will be great for business overall.
Productivity will increase, perhaps dramatically, boosting corporate profits and share prices.
Not everyone thinks this is a good thing. The former Google executive, Dex Hunter-Torricke, is among those warning that the world is heading for disaster. “The most likely outcome is an economy in which corporate profits explode as labour costs fall, while workers’ share of output shrinks,” he wrote recently.
Bad news for workers — but presumably great for investors. In a frequently cited 2023 report, Goldman Sachs analysts suggested that for the top 500 United Stateslisted companies, net margins could increase by four percentage points, or about a quarter, over ten years. Not quite a profits explosion, perhaps. But very helpful for the companies and their shareholders.
There is one big caveat, however.
Productivity improvements should boost profits in the short term. But will higher returns be sustained? Standard economic theory says not.
Companies will cut prices to gain market share and the high profits will attract new entrants into the market, further depressing margins.
And it is not only economic theory that predicts this. So does Jamie Dimon. As head of JP Morgan Chase, the world’s most valuable bank, you might expect him to be pretty bullish about AI. Many experts say the biggest gains from using it will be generated by the largest companies, with the biggest technology budgets and the richest hoards of data. They don’t get much bigger and richer than JP Morgan.
The bank is already spending $2 billion a year on AI but Dimon doesn’t believe this will generate a lasting increase in margins. “This isn’t like you’re going to build three points of margin and you get to keep it — you don’t,” he said last month. Because rival banks are making the same investments, the benefit of the increased efficiency “will eventually be passed on to the customer”, he added.
But when is “eventually”? Even in a world of perfect competition, such gains don’t get competed away immediately. And in many markets, competition is very far from perfect.
Some AI maximalists believe it will be as transformative as the Industrial Revolution. The spread of steam power and mechanisation from the middle of the 18th century delivered a big increase in labour productivity.
Yet real wages did not grow nearly as fast; between 1790 and 1840, industrial wages in Britain stagnated, despite booming business. This so-called Engels Pause, named after Karl Marx’s collaborator Friedrich Engels, resulted in a prolonged increase in profitability. Labour’s share of national income fell and the rate of return on capital doubled.
Subsequent big shifts in technology have had a less clear impact on profits. The boost from the adoption of electricity in the early 20th century was patchier and mostly waned more quickly. The 1990s PC revolution and arrival of the internet was followed by a surge in profits, but economists attribute this more to falling interest rates and growth in trade with India.
Swelling profits coincided with a slowdown in productivity growth in all developed economies over the past 20 years.
It is possible that AI will be different.
Cheerleaders reckon adoption will be so fast and the productivity increase so great that it will drive economic growth on a scale never seen before.
Dario Amodei, co-founder of leading AI firm Anthropic, has talked about annual GDP growth in the US of 10 to 20 per cent. That should certainly turbocharge profits, albeit at the cost of sharply higher unemployment.
But most economists have much more modest forecasts. If they are right, we are back to the question of how sticky the productivity-driven AI boost to profits will prove.
Tera Allas, a former government economist who is a senior adviser at McKinsey, says that in competitive markets, productivity improvements rarely become permanent profits.
“But sustained outperformance for an individual firm is possible when AI strengthens an existing defensible edge — scale, proprietary assets, operating model, customer relationships, or regulatory barriers — rather than AI adoption alone.”
Economists tend to believe that many markets have become less competitive in recent years due partly to increased concentration. This is clearly true in technology. Companies such as Amazon, Apple, Microsoft and Google have been able to exploit their market power to protect huge profits generated by new tech. They will hope to hang on to the gains from AI (as users rather than suppliers).
But in many other industries competition is still fierce, even if concentration has increased. Mike Mayo, a banking analyst at Wells Fargo in New York, says the worry for investors is that the huge amount banks are ploughing into AI may produce higher profits only briefly before becoming simply the cost of staying in the game. In three to five years time it may amount only to “table stakes”, he wrote recently.
All this is very uncertain, but it seems likely that in many sectors much of the AI gains will be quickly lost by shareholders and flow through to customers. Whether those customers will still have jobs is another question.
David Wighton, a former business editor of The Times, is a columnist for Dow Jones
No comments:
Post a Comment