AI investment funds are not all about mega-caps like Alphabet, but about companies in other sectors using AI to increase competition, finds Fiona Rintoul.
If you were to ask artificial intelligence (AI) how to save the planet, it would probably tell you to kill all humans, because they cause the pollution that creates climate change. But that would not be the response you wanted; therefore, you must recalibrate.
“We need to ask the right questions of AI,” says Rani Piputri, head of automated intelligence investing at NN Investment Partners. “In this situation, AI will nudge people to have fewer children.”
This scenario hints at AI’s vast reach. AI affects – or will affect – almost every corner of our lives.
“The advent of the smartphone and its proliferation have really brought down the cost of high-powered computing, and consumers have benefited hugely from that,” says Tom Riley, lead portfolio manager on the Axa WF Framlington Robotech strategy and mandates.
Smartphones have made concepts such as the Internet of Things (IoT), which might sometimes seem hard to grasp, more pervasive. In business, AI’s tentacles are reaching ever further both in terms of sectors and applications, creating opportunities and headaches for fund managers.
“When it comes to AI, people often think of the technology sector and certain enablers such as Google and Microsoft; however, the application of AI and advanced technology can be actually more pronounced and effective in other sectors,” says Ken McAtamney, head of the global equity team at William Blair.
The automation of factories is one example where AI could have powerful implications. For example, AI enables robots to pick difficult-to-identify objects off an assembly line, to inspect parts quickly and accurately and to predict maintenance needs, preventing expensive downtime.
McAtamney also points to the retail sector where big data and advanced computing have enabled companies to apply personalised marketing techniques that offer cheaper products to consumers at opportune moments.
“We are invested in companies that offer these promising applications as well as providing sensors that are the key enablers for collecting data to power AI applications,” he says.
Healthcare is another sector where AI is having a huge – and sometimes underappreciated – impact.
“The amount of technology being incorporated to help drive better patient outcomes is phenomenal,” says Riley.
For example, a company such as Dexcom, which distributes continuous glucose monitoring systems, is revolutionising diabetes management. And robotic surgery, which can reduce scarring, improve recovery times and lower infection risk, is proliferating.
“At the moment, about 2% of surgical procedures are done by robots and around 25% by traditional minimally invasive surgery,” says Riley. “I’d expect that 2% to converge close to that 25% over the coming years.”
AI is now also starting to have a real fundamental impact on company business models, according to Scott Berg, portfolio manager of the T. Rowe Price Global Growth Equity fund. He cites the example of Workday, a cloud-based provider of HR software. Unlike the static software that competitors install on a company’s servers, Workday’s product is constantly learning and its experience with millions of employee records means it can now predict which workers are in danger of growing dissatisfied.
“In our meeting, we had a real sense of the strength and durability of its cashflow,” says Berg. “It was also apparent how powerful its analytics capabilities had become and the improved value-add to the customer.”
Of course, these developments are all hugely disruptive. It is clear there will be losers as well as winners.
“Disruption differs from iteration, which is about improving existing ways of doing things, and from innovation, which is about doing new things,” says Karen Kharmandarian, co-manager of the Thematics Artificial Intelligence & Robotics strategy at Thematics Asset Management.
AI does new things that make the old ones obsolete. The downside for those who cannot keep up has grown disproportionately. This can happen in any part of the economy, says Kharmandarian.
“Avoiding businesses and companies most at risk from AI transforming existing ways of doing business and gaining exposure to those at the forefront that have embraced it successfully and stand to benefit is paramount,” he adds.
The changes wrought by AI sometimes feel almost frightening. For as much as technology has already altered our lives in ways undreamt of even 20 years ago, this is just the beginning.
If AI were a baseball game, we wouldn’t even be in the first innings, says Erik Swords, senior research analyst and portfolio manager at Mellon Investments Corporation. Therefore, it’s an exciting time for technology but also a very uncertain time.
“With all the pros outweighing the cons, I still think there is a lot of work that needs to be done from a regulatory perspective to understand different components,” says Swords. “We always have to be conscious of the unintended consequences of different things that are taking place.”
The importance of trust
Making the most of AI is also a challenge. It’s pointless to collect data for its own sake, but it’s obvious that this does happen. We’ve probably all been shown adverts for socks on the internet – just after buying a pair. Making AI meaningful is harder than just making it.
“It’s much easier to build an AI than to train one,” says Tom Weller, innovation analyst at Evenlode Investment. “AI strategies must be focused on building trust and creating and protecting training data.”
Yes. If AI is going to park your car or diagnose your illness, you clearly need to believe it is as effective or more effective than humans. But how can companies build that trust?
“Training and requisite benchmarking require a trusted community,” says Weller. In medicine, he notes, companies with a huge customer base but no trusted community lost out to slower starters with a stronger trust base – despite an early start.
Customers also need to trust that their data isn’t leaking, or is suitably anonymised, adds Weller. This means synthetic data is an area to watch. Commitment to data analytics is also crucial for a company to make the most of AI’s possibilities.
“Anyone can have an AI strategy, but investors need to truly understand if a company thinks about data analytics as a core competency,” says McAtamney. “An investor needs to determine if management is breaking down data silos and treating analytics as a competitive advantage.”
Trust, then, is perhaps a starting point for filtering potential investments. But how else can fund managers shake out the investment opportunities that this potent but sometimes mind-bending new phase in our economic development is creating?
One issue to consider is market cap. AI opportunities can just as well exist among mega-cap companies as among smaller, earlier-stage firms, but those investors may already be exposed to those mega-cap firms through straightforward global equity funds.
Innovation
A multi-cap approach can mitigate this problem. This is the route that Riley takes at Axa IM, owning Alphabet and Google, but with a bias towards small and mid-cap firms. Ninety percent of the portfolio isn’t in mega-cap, and the biggest names in the portfolio rarely account for more than 4% of it, reducing the risk of double exposure.
“There are a lot of very interesting companies that are focused on smaller emerging areas such as industrial technology and predictive maintenance,” says Riley.
Of course, smaller firms bring risk. That’s where innovation often happens, notes James Cooke, manager of the Ashburton Global Leaders Equity fund, but once past the proof-of-concept phase, new technology requires investment, robust development, distribution and intellectual property-rights management to be commercially successful.
“It often takes a specific management skill set to scale small, successful firms into global contenders,” says Cooke. “For these reasons, many brilliant technology innovations find their way to a relatively small number of mega-cap technology giants.”
Eye-popping returns are available if investors manage to pick the few small technology firms that become hugely successful, but it’s a challenge to identify them. That’s why Axa IM’s Riley normally doesn’t invest in earlier-stage companies but waits for “stability in the business model”, and why Ashburton invests in Alphabet.
“Alphabet generates fantastic returns on capital, and we envisage these will stay substantially above its cost of capital for many years,” says Cooke. “In other words, the company continues to compound intrinsic value – it is a true charging elephant.”
At the same time, there aren’t that many companies that qualify as a pure-play AI vendor, says Swords. “It’s usually something that plugs into something that is already being done and they’re using AI to enhance their existing offering.”
It’s also important to acknowledge that AI does not exist in a vacuum. Berg believes platform companies are best placed to benefit from AI and machine learning over the coming decade but makes the caveat that this may be an area of ongoing conflict between the US and China.
“While AI is an area of structural growth, you are selling in more cyclical end markets,” adds Riley. “One of the benefits of having a broader approach is that it gives you flexibility.”
Flexibility is perhaps the watchword for AI investors. The pace of change is dizzying, and most of us, including regulators, are playing catch-up.
“We’re in a position where technology and the pace of it is accelerating so fast that we don’t have the regulatory oversight – or insight for that matter – to appropriately manage the different puts and takes,” says Swords.
© 2020 funds europe
