As US markets hit fresh records on the back of artificial intelligence-driven optimism, value investors are seeing more attractive deals in other developed markets and emerging economies.  

Brad Kinkelaar, senior managing director and portfolio manager at the Dallas-headquartered Barrow Hanley, said the manager has reduced its US exposure from around 70 per cent of the portfolio two years ago to 55 per cent now.  

“We had been overweight the US for many, many years even though we weren’t invested in the big tech companies that everybody else wanted to own,” Kinkelaar told the Investment Magazine Fiduciary Investors Symposium.  

“We were benefiting from having that US exposure because we had outgrown the rest of the world pretty meaningfully for quite some time.” 

The manager started seeing “attractive value” in European cyclicals, Japan and emerging markets in the first quarter of 2024, according to Kinkelaar.  

“It’s not that we don’t believe that US companies still have more robust fundamentals, but the valuation parameters had shifted pretty dramatically in favour of other markets,” he said. 

“Since we made that shift [to reduce the US markets exposure], those other markets have outperformed on a value basis. It wasn’t an explicit timing bet. It’s just where the value led us, that eventually led to outperformance in other markets as well.” 

Growth investing has outperformed in the five years, driven by the relentless rally of US mega cap stocks, but Kinkelaar said that outperformance can only be maintained if company valuations stay high or continue to go higher.  

“If you have any sense of risk in the marketplace, that is a headwind in a major way. So that’s what we’re trying to manage against – we don’t have to underwrite that risk the same way… as a value investor.” 

One of the top concerns investors have about buying into the AI boom is whether technology companies that have massive capex can translate that into earnings growth in the future.  

Wary investors sent the stock price of Facebook parent company Meta into its biggest daily drop in three years at the end of October as the company announced giant outlays of building AI capabilities.  

Kinkelaar also evoked the case of Baidu – the Google-equivalent in China – which was the first to launch an AI offering in China and has a search assistant, Ernie bot, that claims to have five times the large language parameters as ChatGPT. 

Baidu’s earnings per share growth has remained stagnant over the past several years because while the Ernie bot claims a significant user base – 430 million as of last November – the company has difficulty monetising it.  

And these won’t be standalone cases in the AI gold rush, Kinkelaar said. 

“So far, the cash flow spenders have been rewarded just like the cash flow recipients. And I think at some point that has to change,” he said. 

“Monetising AI so far is an unproven bet – it’s an unproven speculation. We’ve been able to benefit by getting in front of what appears to be fairly predictable cash flows over the next few years, but there’s a huge, massive bet out there on who’s going to be the winner with the AI technology – not just the infrastructure – and that’s really unknown.” 

Kinkelaar warned that a big proportion of the AI infrastructure is “spec building”, where the infrastructure was established before securing a tenant.  

“Sam Altman said OpenAI has a ‘trillion-dollar’ commitment [to invest in AI infrastructure] – and I put air quotes around that – and he’s saying he needs US$10 trillion of commitments by 2033 [to build 250 gigawatts of computing capacity]. This is all spec building based on debt financing,” he said. 

“At some point we’ve got to recognise that that some of this starts to smell like the build out for the internet [in the dot come bubble]. 

“Maybe AI ends up being a common good that we all benefit from, but there’s only a few players that can really make money on it.”