Today’s Special Report comes courtesy of Cannaccord’s Justin Oliver
“The 9000 series is the most reliable computer ever made. No 9000 computer has ever made a mistake or distorted information. We are all, by any practical definition of the words, fool proof and incapable of error.”
HAL – 2001: A Space Odyssey
While we are still quite some way from true ‘artificial intelligence’ (AI) as science fiction writers might understand it, the term is being used with ever greater frequency. In investment terms, the concept stretches beyond laboratories of scientists trying to develop neural networks. It now includes machine learning (a field of computer science where computers learn without being programmed), deep data (looking for specific information to help predict trends or make calculations), natural language and autonomous vehicles. It also includes any company which could feasibly benefit from these technological advances. These areas are presented hand in hand with industrial automation and smart robotics to present a holistic investment universe.
To help investors capitalise on the march of technology, investment providers have developed several ways to access this theme. At least three actively managed funds have been launched in the UK in the past 12 months, which all have some degree of AI incorporated as a theme, while there are also exchange traded funds (ETFs) available which offer exposure to the specific areas of automation and robotics.
The London-listed Robo Global Robotics and Automation GO UCITS ETF share issuance has risen from just 4.4 million shares in May 2016 to 57.9 million by the middle of November 2017. iShares Automation & Robotics was launched in September 2016, but already has a market capitalisation of US$1.3bn.
So should we, as investors, be seeking to capitalise on this trend?
A compelling investment opportunity
At first glance, the economics of the AI and robotics industry appear compelling. According to a prominent AI fund manager, “we are at the start of the adoption of AI. Those positioned to benefit will enjoy a sustained period of growth.” Even more eye catchingly from an economic perspective, “AI has the potential to double economic growth rates over the next 20 years.”
The corporate world seems to agree. CB Insights, which offers “machine learning, algorithms and data visualisation to help replace the 3 Gs (Google searches, gut instinct and guys with MBAs)” to predict future trends, estimates that US$1.5bn has been invested in AI companies in the last five years. Many of the largest companies in the world are investing in AI and machine learning, including Apple, Google, Amazon, IBM, Intel, Microsoft, Nvidia, Qualcomm and Tesla.
Research and Markets, an organisation with 1,700 research teams in 81 countries, expects the global AI market to reach US$36bn of revenue by 2025, growing at a compound annual growth rate of 57% between 2017 to 2025. Meanwhile, McKinsey estimates that 30% of tasks in 60% of occupations could be automated.
Robotics also appears to be a growth industry. The International Federation of Robotics highlights that in 2016 global sales of robots increased by 18%, to US$13.1bn, while robot supplies are expected to rise by 21% in Asia, 16% in the US and 8% in Europe.
Unsurprisingly, the shares of companies involved in automation have responded favourably. Fanuc rose nearly 35% between mid-September and mid-November 2017, while Yaskawa and Kuka doubled within the year.
What’s not to like?
At CGWM we prefer to avoid investing in fads and fashions. While AI may not prove to be a passing fad, it is certainly in fashion at present.
There is also a fallacy that faster growth automatically feeds through to higher investment returns. Unfortunately, there are flaws in this argument. For example, many of the growth statistics surrounding AI focus on revenues, as profitability within much of the industry is a scarce commodity. This is reminiscent of the late 1990s tech boom, and at this stage it is impossible to know which areas of AI, and which companies, are likely to survive and thrive. If history is any guide, there will be as many failures as successes.
We should beware of relying heavily on forecasts which look too far into the future, particularly from those who have a vested interest in the area. While economic growth may be boosted by the adoption of AI and deep-thinking algorithms, this is by no means certain. Just as today’s world looks nothing like the one envisaged by sci-fi writers of the past, tomorrow’s world may look vastly different from current predictions. The forecast ‘winners’ in certain scenarios may not be the existing incumbents; they may not even exist yet.
History also shows that the benefits of new technologies often take longer to materialise than first expected. For example, Paul David, the academic economist who has undertaken analysis of scientific progress and technological changes, found that when electricity was first adopted within factories, they became slightly more productive. It was only later, when the factories went further and began changing their configurations to capitalise on electricity, that the surge in productivity really began. Maybe humans will need to adapt their behaviour to deliver the true benefits of AI and automation.
Another issue is whether investors can truly allocate to or between the various AI sectors, or whether a wider technology exposure should prove sufficient. Would a broad technology fund, managed by experts in their field and who understand the fact and fiction of the AI investment universe, be a better investment route? The definitions of AI are so variable, and the investment universe so ‘non-standard’, that a compelling top-down view can quickly become blurred.
We always need to be sceptical of the claims made by the proponents of investment opportunities. The prospects offered by AI, robotics and automation certainly seem compelling. However, we don’t take such claims at face value, or assume that they automatically guarantee a successful investment outcome. Instead, we will weigh the pros and cons of the investment rationale and suggest that investors also tread carefully before committing to this area.