After the buzz, investors are doing their own homework on AI

The rapid adoption of generative artificial intelligence has boosted markets this year, but after the initial euphoria, investors are waking up to the possible risks, including the need to be highly selective in stock-picking.

Businesses ranging from IT services and consulting to media, information and education are now under portfolio managers' microscopes to assess the potential for AI disruption.

The overall impact for corporate profitability is seen as hugely positive. Yet beyond Nvidia and other obvious winners in the chip sector, analysts warn there might also be losers across Europe and the US.

McKinsey says generative AI could add US$7.3 trillion (about RM34.2 trillion) in value to the world economy each year and believes half of today's work activities could be automated between 2030 and 2060.

That, however, means corporates also face big challenges, like redundancies and rethinking their business models, if they want to fully realise AI's potential.

"It's not a given that AI will only have a positive impact. There could be a deflationary effect," said Gilles Guibout, who helps manage over US$900.44 billion as head of European equities at AXA Investment Managers in Paris.

In some cases, clients could negotiate price cuts, he said, while staff-light newcomers could erode existing players' market share while they are busy redesigning their processes.

That might reduce sales growth and cause share price underperformance, especially for companies that face strong competition or where growth depends on headcount.

"Take IT services: if one hundred people are no longer needed for coding, but only half or a third of that, customers will be asking for lower prices," said Guibout.

The latest Bank of America survey in June showed 29% of global investors don't expect AI to increase profits or jobs. That compares to 40% that do expect a boost.

AI not always 'good'

Concerns about AI have already manifested across markets.

Shares in companies like French outsourcing firm Teleperformance and US-based Taskus, which manage call centres and other services seen as vulnerable to being replaced by bots, have both lost around 30% this year.

In education, UK's Pearson slumped 15% one day in May after US peer Chegg, down 62% this year, said significant student interest for the Microsoft-backed ChatGPT bot was hitting customer growth.

A few days later, Pearson held a call to explain its AI strategy, a sign of growing interest among investors to go deeper into how corporates are dealing with the transition.

Teleperformance, which employs 410,000 staff in 170 countries, held its AI investor day on Wednesday.

Some analysts say price falls have been excessive in certain cases, exaggerating the concerns over earnings growth.

"There's a lot of focus on the risks that generative AI can bring. This has ultimately become a bit overdone," Thomas McGarrity, head of equities at RBC Wealth Management.

He sounded confident over the capacity of some professional information and data providers, which own proprietary data, to integrate generative AI into their products.

Others, meanwhile, remain cautious, saying the fast adoption of cheaper AI-powered offerings could slow growth as soon as order backlogs of more conventional services are fulfilled.

Andrea Scauri, portfolio manager at Lemanik, said uncertainty over AI has deterred him from investing in some IT services stocks, despite valuations looking attractive.

On the other hand, Scauri said he sees larger players like Accenture as better equipped to navigate the transition and deploy necessary capex.

Accenture unveiled a US$3-billion investment plan to power its AI efforts this month, three months after announcing 19,000 layoffs, or about 2.5% of its workforce.

Its shares have risen 19% this year and French peer Capgemini CAPP.PA is up 13%. Firms such as Relx which handle regulated information, are also seen as less exposed to potential AI headwinds.

Cristina Matti, small and midcaps portfolio manager at Amundi, said indiscriminate investing was not an option for investors seeking AI exposure.

"Do not buy just for the sake of gaining exposure. It’s important to do your homework," she said.