Last month, US-based computer chipmaker Nvidia Corp surprised markets with sales guidance into 2024 that was much higher than expected, pushing its market capitalisation to US$1 trillion ($1.34 trillion), just behind the likes of Amazon and ahead of Meta.
Nvidia, which produces highly sophisticated chips central to building AI models such as ChatGPT, has more than tripled since October 2022, far outpacing any other member of the broad-market S&P 500 index.
This breakthrough for Nvidia marks the beginning of the AI boom that analysts have been speculating about since 2021. Fortune Business Insights had estimated that the global AI market size would grow at a CAGR of 40% to reach US$360.4 billion by 2028 before ChatGTP made waves. Today, it is projected to grow to US$2.025 trillion by 2030 from US$515.31 billion this year.
The growth trajectory is tantalising for any onlooker and businesses and investors alike will quickly ask themselves how they can cash in on this boom. Asset managers observing the space for almost a decade say that Nvidia’s explosion was years in the making.
“The real-world implementation of AI has appeared over the last 10 to 15 years in various internet products for recommending the next best story, for recognising content and pictures, and turning conversations into transcriptions,” says Jonathan Curtis, director of portfolio management of Franklin Equity Group.
See also: Unveiling value opportunities in energy, healthcare and technology
Investors are “really good at overestimating the near-term impact of technology, and not so good at understanding the long-term impact”, says Jonathan Curtis of Franklin Equity Group / Photo: Franklin Templeton
But what has changed since November 2022 is the advent of accessible generative AI models that can generate human-like responses. This is a product of many iterations of semiconductor chips and hyper scalers (large cloud service providers), which has allowed large language models to come to fruition, says Anjali Bastianpillai, a senior client portfolio manager at Pictet Asset Management.
These developments mark a significant change. “AI has been advancing steadily through the highs and lows of the last decade and I think the stage is now finally set for game-changing AI use cases,” adds Stephanie Leung, chief investment officer of StashAway.
See also: Time to rethink traditional thinking in emerging markets
So how can investors catch this AI boom? The asset managers who spoke with The Edge Singapore said that the growth of AI is “just at the beginning of a long trend” and while at present it might seem like some clear winners have emerged, the industry is only sitting on the cusp of exponential growth.
It is tempting to make big bets on how trends will play out but staying invested in a diversified portfolio for the long term is a winning strategy for AI, says Stephanie Leung of StashAway / Photo: StashAway
Winning strategy to cash in on the AI boom
The excitement over AI has made it tempting to bet on Nvidia, with its share price surging to nearly 65x forward FY2024 earnings at its peak in late May. Yet, betting heavily on a single company comes with its risks. “If you’re waiting for the ‘next drop’ before entering the market, that could mean missing out on growth opportunities which could prove costly in the long run,” says Leung of StashAway.
Matthew Cioppa from Franklin Equity Group says that even though Nvidia is a clear leader today from a revenue concentration perspective, there are “plenty of others” in the semiconductor businesses that stand to benefit from this AI trend.
Nvidia is a clear leader today, but there are “plenty of others” in the semiconductor businesses that stand to benefit from this AI trend, says Matthew Cioppa of Franklin Equity Group / Photo: Franklin Templeton
For more stories about where money flows, click here for Capital Section
Cioppa, who is a research analyst, advises looking at other firms deploying AI in different or adjacent sectors, such as Qualcomm and Broadcom. Marvell Technology, for one, is worth looking at as it provides ultrafast interconnectivity between Nvidia’s graphics processing units (GPU) in a data centre. He also names Lattice Semiconductor, which provides low-power programmable chips that can power smart cities or smart homes, a company he believes can stand to benefit.
The analyst believes that Nvidia will continue to hold a leading market share in GPUs for a long time but the recent introduction of a new GPU by Advanced Micro Devices (AMD) could position the American company as a second supplier of GPUs.
The next layer of companies to look at would be cloud companies, says Curtis of Franklin Equity Group. “Google, Microsoft, Amazon and even Oracle. Microsoft has a close partnership with OpenAI, one of the best builders of these models … pay attention to what they’re saying,” he says. Since 2019, Microsoft has invested a cumulative total of US$13 billion in OpenAI, lifting the latter’s valuation to around US$29 billion.
But it is also crucial for investors to become “experts” in the space. “Do your fundamental work and look for the signals of quality businesses that can generate sustainable demand and ultimately are trading at a discount to their intrinsic value,” adds Curtis.
Citing Oracle’s US$2 billion worth of committed cloud businesses with future AI partners as an example, Curtis encourages investors to “get into the weeds ... get into the details and pay attention to those numbers and see if they’re continuing to move up”.
Finally, the asset managers from Franklin Equity Group say investors should watch if everyday customers are adopting these innovations, whether adoption rates are moving or if customers are paying for the different suites of office productivity capabilities.
The adoption of new technology by everyday customers will determine whether a company has a good business case or ideas on how to properly monetise the software, adds Pictet’s Bastianpillai. However, she warns that the world is only at the beginning of a long trend in investing in AI, which makes it difficult to figure out who the real winners are.
“The difficulty today is in two things, the first is who is going to monetise this in the best possible way, the second is to look for businesses that can actually improve efficiencies, reduce costs and increase productivity,” Bastianpillai adds.
Look for businesses using AI that can monetise, improve efficiencies, reduce costs and increase productivity, says Anjali Bastianpillai of Pictet Asset Management / Photo: Pictet
Diversification is the only free lunch in finance
But is this AI boom just a bubble? After all, the tech industry, more than any others, is notorious for trumpeting the next big thing only for the hype to vaporise into oblivion when something fancier comes around. According to Curtis, investors are “really good at overestimating the near-term impact of technology and not so good at understanding the long-term impact”, which might create a little bubble right here and now.
As seen with Nvidia, Curtis believes investors who underestimated the impact on Nvidia’s business were later shocked by the positive outlook that it painted. As a result, every technology company today is experimenting with how to bring generative AI into their tools, regardless of whether there is any actual demand or a good business model, Curtis adds.
“We’re seeing a lot of revenue happening in chips, we’re seeing a lot of revenue happening in cloud services, where these models are being built and supported. But we haven’t seen whether the dogs are going to eat the dog food yet, where their end-users are going to use this technology and pay for it,” says Curtis.
For asset managers like Curtis and Cioppa, there has been an increase in conversations with companies and investors to gain a better understanding of the developments. “Every question is almost about AI — how are you using it, what are you experimenting with and what are you learning about your business models?” says Curtis.
Curtis cautions against getting too wrapped up in how a company might perform in the short term and advises to think about the long-term opportunities presented. “When investors are overexcited, be more cautious, and when they’re under-excited be more greedy,” he adds.
Leung believes that the broader market is not in a bubble and is trading at reasonable valuations. She notes that the forward P/E for the tech-heavy Nasdaq index has run up to 28x as of mid-June from a trough of about 20x last year. “That’s still below a recent peak of about 32x in 2021 and well below a P/E of 75x during the height of the dot-com bubble. The figure for the S&P 500, meanwhile, has risen to around 20x — not too far from its long-term average of around 18x,” she reasons.
Nonetheless, Leung too advises investors to diversify their portfolios to avoid concentration risks while maintaining a certain level of exposure to AI. After all, even the best fund managers in the world are wrong more than 40% of the time, she says.
“It can be tempting to make big bets on how these trends will play out but staying invested in a diversified portfolio for the long term can still be a winning strategy for AI,” says Leung, who suggests exposure to related sectors such as robotics, semiconductors and cloud computing, along with “balancing assets”, like gold, can help protect against downside risk.
