Our Company and Sites
June 2018

Investing in breakthroughs: Robotics and artificial intelligence

Christopher Dhanraj
Director
Head of ETF Investment Strategy

Robotics and artificial intelligence (AI) are at the forefront of technological advances that represent a potential transformational “megatrend,” a new industrial revolution. These technologies are already rapidly impacting many industries and occupations, while in some sectors the change is just beginning. Investors may want to consider the long-term potential of these
technological breakthroughs.

Key points:

  • Robotics and artificial intelligence represent a long-term transformational shift, affecting the way businesses operate – and how people live. Investors may be able to benefit from these changes for the long term.
  • Robotics already has long-established industrial applications, but advancements in the field are expanding into surgical and consumer applications.
  • AI has significant potential across various sectors, including industrials, financials and healthcare and others.
  • Investors have poured over $20 billion into robotic and AI-focused funds in the past three years.
  • Risks to the space include organizations and regulation not keeping pace with change, and excessive investor excitement.

Setting the scene: the fourth
industrial revolution

The rise of robotics and AI is what we define as a megatrend, a structural shift that is fundamentally transforming the global economy in a way that could have significant repercussions for the long term. The mass adoption of robots, automation and artificial AI could represent a fourth industrial revolution after steam, mass production and electronics, according to the World Economic Forum.1

Robotics

Robots have found a place in the global economy for a number of years thanks to industrial applications, and the demand does not appear to be abating (see Figure 1). The automotive and electronics industries in particular have continued to support an increase in the demand for robots over the last three years (see Figure 2), particularly to support manufacturing processes.

Estimates of the value of the robotics industry vary, but one report by IDC predicted that worldwide spending on robotics and related services will amount to $188 billion in 2020 – more than double the $91.5 billion figure for 2016.2

Figure 1: Estimated annual worldwide
supply of industrial robots

2008-2016 and 2017-2020*

Estimated annual worldwide supply of industrial robots

Source: International Federation of Robotics (IFR), World Robotics Report 2017, September 2017.
* Represents forecasts by the IFR quoted in the World Robotics Report 2017.
There is no guarantee that the forecasts will come to pass.

Figure 2: Continued increase in major industries

Estimated annual supply of industrial robots at year-end by
industries worldwide 2014-2016

Estimated annual supply of industrial robots at year-end by industries worldwide

Source: International Federation of Robotics, World Robotics Report 2017, September 2017.

Robotic uses continue to expand. For example, one application with significant growth expected on the horizon is the field of surgical robotics. In addition, robots have also found their way into households as the technology gets cheaper and smarter. The worldwide number of domestic household robots is set to increase to 31 million between 2016 and 2019 (see Figure 3).

Figure 3: Service robots for personal/domestic use

Unit sales forecast 2016-2019*, 2015 and 2014 & 2016

Service robots for personal/domestic use

Source: International Federation of Robotics (IFR), World Robotics Report 2017, September 2017.
* Represents forecasts by the IFR quoted in the World Robotics Report 2017.
There is no guarantee that the forecasts will come to pass.

Artificial intelligence

Artificial intelligence is arguably more transformative than robotics. AI encompasses the development of computers able to learn to perform complex tasks, including some that are beyond human intelligence.3 Faster computers and exponentially-expanding data have allowed advances in fields such as deep learning, which demonstrate that it can be both practical and profitable to put AI to work.4

It has been estimated that AI industry revenues will grow to $127 billion by 2025, from $2.1bn in 2015, implying a compound annual growth rate (CAGR) of 51%.5 AI’s impact on the broader economy is hard to quantify, though PwC estimates that it could contribute $15.7 trillion to the global economy in 2030 – more than the current output of India and China combined – through $6.6 trillion of productivity impact and $9.1 trillion of consumption side effects.6 Such estimates are naturally subject to a huge degree of inherent uncertainty, but the tone of analyst expectations is optimistic.

Applications of AI include a range of uses:

  • Electronic devices that consumers can use to control their homes.
  • Retailers using predictive analytics to better model and predict buyer behavior, as well as provide customer service.
  • Manufacturers using machine vision to inspect assembly lines and guide robots for quality control.
  • Healthcare providers using AI to identify patterns in large unstructured datasets to aid medical diagnoses, as well as support in patient management.
  • Financial advisers using AI to provide affordable advice to a broader range of clients, and banks using AI to improve fraud detection and regulatory compliance.

The improvement in the technology facilitating these applications is impressive. For example, in 2010, Google’s Machine Learning speech recognition accuracy was about 70%7, but by 2017 it achieved an error rate of less than 5%, better than humans.8 Machines have also been able to overtake human performance in the case of computer vision, as demonstrated by the results of a competition involving machine learning researchers submitting algorithms for the detection of objects (see Figure 4). The accuracy of the best-performing algorithms now exceeds the accuracy level expected of a human performing the same task.9

Figure 4: The ability of AI systems to recognize objects has improved markedly to the point where the best systems now exceed human performance

The ability of AI systems to recognize objects has imporved makedly to the point where the best systems now exceed human performance

Source: McKinsey Global Institute, April 2018, Notes from the AI frontier: insights from hundreds of use cases, referring to Olga Russakovsky et al., “ImageNet Large Scale Visual Recognition Challenge,” International Journal of Computer Vision, volume 115, issue 3, December 2015.

It has been forecast that AI will be the largest driver of tech spending over the next decade.10 Internal investment by large corporations in AI was estimated to be between $18 billion - $27 billion in 2016 alone, while external investment (from venture capital, private equity, etc.) accounted for $8 billion – $12 billion.11

Investment implications,
flows and risks

Comparing the time spent by companies discussing AI in their earnings calls reveals an interesting validation of AI’s potential as judged by the market. Companies who spent the most time discussing artificial intelligence in their investor earnings conference calls have started to outperform those who spent the least time on it, regardless of sector. However, this has only occurred since 2017, in line with the more recent boom of interest in the space and perhaps, hopes for its potential
(see Figure 5).

Figure 5: Rolling return differential for companies who mention artificial intelligence most vs. least

Rolling return differential for companies who mention artificial intelligence most vs. least

Source: BlackRock Investment Institute and BlackRock Multi-Asset Strategies, with data from Conference Call corpus and MSCI, as of May 2018. The line shows the rolling two-year sector-neutral return differential between the top 50% of global developed market companies (included in the MSCI World Index) mentioning AI on conference calls and the bottom 50%. The top and bottom 50% were based on time spent discussing AI. Index performance is for illustrative purposes only. Index performance does not reflect any management fees, transaction costs or expenses. Indexes are unmanaged and one cannot invest directly in an index. Past performance does not guarantee future results.

Investor interest in both robotics and AI is also witnessed by flows into listed and unlisted vehicles providing exposure to the theme. Assets under management for robotics and AI funds are now estimated to be over $25 billion, a striking increase from a mere $59.8 million in a handful of funds in 2013.12

Investing in a newly developing megatrend does come with unique risks. First, investor excitement always has the potential to send valuations ahead of fundamentals, as has occurred in certain unlisted markets such as venture capital. In addition, the progress of technology may be limited by the ability of firms and regulators to keep up with the pace of change, and grapple with privacy issues.

The bottom line

Robotics and artificial intelligence have the potential to disrupt entire industries, transform economies and reward investors seeking growth. The potential technological breakthroughs, applications and adoption of the technologies are still in the early stages. We believe that investors should consider the long-term benefits of this still-developing space, both by accessing companies that are embracing robotics and AI and leading the way, and those that are enabling their disruptive growth.