Market Musings 270523: Is AI the next Big Thing (for investors)?

Market Musings 270523:
Is AI the next Big Thing (for investors)?

Nvidia’s post results surge underlines the hype over ChatGPT and AI

Summary:

• Artificial Intelligence: A Bubble, or the Next Big Thing?

• Gartner’s Hype Cycle

• ChatGPT the trigger for AI excitement

• Nvidia the Poster Child for Investing in AI

• If you want to invest in AI, how else can you invest apart from Nvidia?

• Finally, what could the macro impact of AI really be?

Weekly Podcast:
The US debt ceiling: the final countdown

• Is 1 June a scary date?

• Can the American President really invoke the 14th amendment?

• Why are CDS prices so high but the probability of the US default is not?

• Why is the T-bill dislocated?

Artificial Intelligence: A Bubble, or the Next Big Thing?

There is no doubt that there is a huge amount of hype in today’s financial markets around the theme of Artificial intelligence (AI).

While AI has been around for many years (I myself completed a Doctorate in Artificial Intelligence in the early 1990s), the technology finally seems to be hitting the mainstream in terms of applications in the everyday and business worlds.

For many in Main Street, the image of Artificial Intelligence is perhaps best personified by the risk that one day, computers will rule over people, Terminator-style.

However, I would rather argue that Artificial Intelligence and Machine Learning are technologies that can potentially magnify once again the impact of Information Technology on our everyday lives, after the advent of the Internet/World Wide Web and then of Smartphones.

Of course, as Artificial Intelligence captures the imagination of investors in much the same way that the Metaverse did a couple of years or so ago, there is always the risk that investors today will price in over-optimistic sales and earnings expectations. This is a hype pattern that repeats itself over and over again, as it has done for over a hundred years in stock markets with former “technologies” such as railroads, radio and mainframe computers.

I am not saying that there is no solid reasoning behind this excitement. For instance, consultancy PwC sees AI-related productivity savings and investments generating $15.7 trillion worth of global economic output by 2030, almost equivalent to the gross domestic product of China.

But I would temper my excitement, for historical reasons…

Gartner’s Hype Cycle

The advent of new technologies tends to follow the Hype Cycle mapped out by consultancy Gartner.

According to Gartner,

The Hype Cycle is a graphical depiction of a common pattern that arises with each new technology or other innovation. Although many of Gartner’s Hype Cycles focus on specific technologies or innovations, the same pattern of hype and disillusionment applies to higher-level concepts such as IT methodologies and management disciplines.

Hype Cycles characterize the typical progression of innovation, from overenthusiasm through a period of disillusionment to an eventual understanding of the innovation’s relevance and role in a market or domain.

Gartner’s Hype Cycle

Source: Gartner

This hype cycle can be then closely linked for investors to asset bubbles, which have occurred many times in the past with new and hyped technologies such as cryptocurrencies, the Internet and radio.

The Psychology of Asset Price Bubbles

Source: Dr. Jean-Paul Rodrigue

This is most certainly a psychological phenomenon and not a question of intelligence - in the past, there have been plenty of examples of famous intellectuals being sucked into asset bubbles, such as Sir Isaac Newton in the South Sea Bubble of 1720. Greed and the madness of crowds can get the better of even the most intelligent amongst us…

Examples of Asset Bubbles in History

Source: Investopedia

Conclusion: Investors should beware the hype around the “next great thing”, as it can often mean investing in an asset bubble when asset valuations are already heavily inflated, running the risk of taking a huge loss when the asset bubble finally bursts.

ChatGPT the trigger for AI excitement

The trigger for this latest excitement around AI comes from the mainstream adoption of the use of ChatGPT, which holds much promise in a number of different types of business.

Reuters reported that AI bot ChatGPT reached an estimated 100 million active monthly users last month, a mere two months from launch, making it the "fastest-growing consumer application in history" according to UBS.

ChatGPT outstrips adoption speed of all previous technologies

Technology is adopted ever faster in today’s world

Source: Statista

Nvidia the Poster Child for Investing in AI

To be frank, we are too early in the adoption cycle of Artificial Intelligence techniques and programs to know who the real long-term corporate and business winners will be.

I have long advocated a “picks and shovels” approach to investing in AI for this reason - preferring to invest in sectors and companies that will provide the hardware and software tools necessary for the widespread adoption of AI and machine learning in large corporations. I have thus advocated investment in both semiconductors and cybersecurity subsectors of Information Technology.

It seems that investors are going one further in this approach in identifying Nvidia as the key long-term winner of the boom in AI applications - given that their Graphic Processing Units are well-placed to provide the heavy processing power necessary for AI models to “learn” by training on large datasets.

Nvidia is the poster child for the AI mini-bubble

Nividia’s Graphic Processing Units GPUs) are the AI processor of choice

Source: tradingview.com

Nvidia is now richly priced following a 28% post-results surge since Wednesday last week, valued at 35 times sales.

At this point it is also worth remembering that a great company is not necessarily a great investment.

So even if we accept that Nvidia is an exceptional company that is very well-placed to profit from growth in AI investment, we then have to try to figure out how much of this long-term growth is already captured in today’s valuation. Given how high Nvidia’s valuation has risen of late, the temptation is to say “a lot”.

If you want to invest in AI, how else can you invest apart from Nvidia?

I maintain that it is simply too early to be able to reliably identify the long-term winners from any supercharged growth trend in AI. The eventual winner(s) may not even be listed in the public stock markets today…

A slightly more diversified approach might be to invest in funds and ETFs focused on Industrial Automation and Robotics, as this is an area which should benefit from leverage to AI and also from other megatrends such as ageing populations, nearshoring of Asian goods production and the need for companies to reduce their exposure to high wage growth.

The iShares RBOT Automation & Robotics UCITS ETF is one such investing vehicle, which has performed pretty well since the beginning of 2020.

This contains exposure not only to Nvidia (the largest holding, but only a 2.25% weight) but also to a number of leading industrial automation and robotics companies based in the US (50% fund exposure), Japan (17% exposure) and Germany (6%).

RBOT iShares Auitomation & Robotics ETF (in USD)

RBOT ETF has outstripped global stocks over the last few years

Source: tradingview.com

Finally, what could the macro impact of AI really be?

The most significant impact of AI in the real world may be much more about “creative destruction” than benefit for certain mega-cap technology companies like Nvidia, Google or Microsoft.

Rather, the widespread application of AI and chatbots could mean the devaluation of a whole swathe of white collar office jobs which involve a lot of routine, repetitive tasks. Think of call centres, or junior lawyers drawing up and checking standard business contracts, or junior auditors for instance.

Just as previous waves of industrial innovation have allowed for the mass production of goods and for people to move away from working in agriculture as machinery and fertiliser use massively improved agricultural productivity over the decades, so a similar phenomenon may be seen in a number of service industries. We are already seeing the beginnings of the introduction of this trend in automatic voice recognition in call centres to better direct phone enquiries and even to answer simple queries without human intervention. We also see this in the widespread introduction of automated self check-out tills in supermarkets, which require a minimum of supervisory staff to handle problems and watch for fraud, as opposed to one cashier per manual till.

Ultimately, companies may be forced to adopt these technologies in order to simply remain competitive, if others in their industry are adopting AI and improving productivity and cutting costs are a result. Recall the story of Blockbuster, which did not change its business model quickly enough with the advent of Netflix video streaming, and went out of business as a result. Or the example of Nokia, which in the pre-iPhone era had a global smartphone market share of over 50% in early 2007. Today, it is zero, as in 2007 it decided to retain its proprietary Symbian operating system and did not adapt quickly enough to the rise of the iPhone with its iOS operating system, and the Android operating system.

Today economists worry about the surge in wage growth that we are experiencing in the US and Europe at present, at a time when unemployment rates are at or close to historic lows and when long-term illness and incapacity accounts for more and more of the working-age population. But if both manufacturing and service-sector companies really implement AI and machine learning software in their business models to its fullest potential over the next few years, then we could see the elimination of huge numbers of jobs as they exist today, potentially creating a wave of deflation in wages in these industries.

Of course, this technology will also lead to the creation of new types of jobs which could be based more on human creativity for instance, which AI may struggle to really replicate in the near term.

There is huge potential to improve the effectiveness of new drug discovery and of diagnostics in Healthcare, and also to accelerate the democratisation of affordable higher education and of learning of essential skills via AI-based teaching platforms. These are just two examples of industries that could see a huge acceleration in productivity as a result of AI technologies and techniques.

Have a great week

Edmund