How to move on
While a few shares have dominated the AI bull market, the sectors that may be the main beneficiaries of AI application have generally not performed well, often because they have not been “AI enough”. For a while, Apple seemed lukewarm on the subject and then, in mid-June, it gave the analysts all the AI they wanted to hear and their shares jumped by 10 per cent.
Some of us might be rather wary of our smartphones doing even more with our personal data and suspect there is an element of Siri about this, which is fine for those who want it but not for all.
Companies that have emphasised the slow build of AI revenues have seen their shares fall, but may be building a more solid business longer term.
Consultancy Accenture tried to explain that the AI solutions their clients need are in development, and saw its share price fall by more than 20 per cent, despite the fact that the range of corporate tech services it will offer in future has clearly grown significantly with AI.
Cybersecurity group Palo Alto Networks used the AI boom to sign up customers to multi-year subscription packages, swapping short-term income for longer-term income growth, and saw its shares fall by around 15 per cent.
Longer-term AI valuations need to be supported by the amount people will pay for the extra functionality and features. In the consumer space, apart from a different sort of internet search result available on Bing and others, there are few new features to evaluate.
AI is already well established in some business areas, and the productivity improvement it brings will bring significant longer-term value to the tech companies that enable it. However, some of the productivity gain will be captured by the companies themselves.
A good example of this is healthcare. If you have copious data on operations, drug regimens and changes in patient health over decades, you can improve outcomes while also identifying ineffective and wasteful practices. The largest US health insurance companies have been gathering and processing such data for years.
Fairly small changes in their medical expense ratios — the number of health claims in a year — can produce a large change in shareholder returns. The best of these companies, UnitedHealth, has seen its shares perform poorly against a market fuelled by AI. But what would the shares do if the group suggested it was benefiting from it?
Doubtless many more companies will include an AI claim in their plans over the next year. Some of these claims will be valid and many will be exaggerations. Some claims will be laughable. I imagine investor relations teams are busy trying to put AI references in a range of chief executive statements for the September results season.
