Beware the AI stock bubble… and how you can still invest in the AI revolution safely

The cheap money era of ultra-low interest rates and quantitative easing has ended. The Federal Reserve raised federal fund rates to 4.50%-4.75%, a level last witnessed in October 2007. Despite the conclusion of an epoch, speculative bubbles continue to emerge. Market frothiness seems to be the theme in recent years.

First there was the GameStop (NYSE: GME) rally that subsequently crashed in March 2021, during which the stock lost 40% of its value within half an hour. Then came the crypto winter marked by key events such as the stablecoin terraUSD sell-off and FTX bankruptcy. In conjunction, another digital asset bubble was in the making. The non-fungible tokens (NFT) bubble showed signs of bursting as prices dipped and trade volumes plummeted. Today, the hype around artificial intelligence (AI) swells as another bubble shows signs of forming yet again.

Soaring AI stock prices

The stock market registered positive returns in 2023. On a year-to-date basis, NASDAQ, NYSE, and the S&P500 returned 9.9%, 2.3%, and 3.9% respectively. These numbers nevertheless pale in comparison to pureplay AI stocks. One of such stock is (NYSE: AI) which uses AI to help businesses in areas of supply chain management and customer service. Another is SoundHound AI (NASDAQ: SOUN) which develops voice assistants for use in the automotive and consumer service sectors. Both stocks at least doubled since the beginning of this year. Despite so, neither can hold a candle to (NYSE: BBAI), a company that predominantly serves the defence industry using its predictive analytic. This stock rallied by over 400% in 2023.

Hype is fuelled by consumers and investors

When share prices climb steeply, it is usually a result of the market pricing in high expected growth. This bullish run was sparked by ChatGPT, a chatbot developed by OpenAI which is in turn backed by Microsoft (NASDAQ: MSFT). Soon after, tech heavyweights like Google’s parent, Alphabet (NASDAQ: GOOG) unveiled its chatbot Bard in hopes to rival ChatGPT. The issue here is that consumers expect growth way too soon. According to UBS, it took Instagram two and a half years to reach 100 million users and TikTok nine months after its global launch. ChatGPT only took two months. The rapid ballooning of users, unfortunately, is merely the first sign of irrational exuberance.

Besides consumers, investors have also priced in the precipitous growth of AI prematurely. All three pureplay AI stocks are loss-making. More specifically, and SoundHound AI have net losses that dwarf their revenues, mainly due to high research and development expenses. Furthermore, the management of these three companies have projected persistent losses going forward. Even, which was once barely profitable, affirmed its guidance of negative earnings before interest, tax, depreciation, and amortisation (EBITDA) for the second half of 2022.

Investors seem to therefore be content only with a growing market size and revenue, with little regard for profitability. This is alarming primarily because the solvency of a company is dependent on its profitability. Recurring losses chip away at a company’s cash reserve, which is used for maintaining operations, fuelling growth, and repaying its lenders. This means companies that accumulate losses are at a higher risk of facing illiquidity and possibly, insolvency. Hence, jacking up share prices across the AI board despite the lack of a clear runway to profitability is the second sign of a bubble.

AI advancement is a limiting factor

AI is a very broad umbrella term. According to McKinsey, it refers to the ‘ability of a machine to perform cognitive functions typically associated with human minds’. The recent hype pertains to a specific use case of AI in chatbots. One reason why investors and consumers jump onboard the bandwagon is the disillusioned belief or hope that an AI can think like a human in the very near future – a concept known as artificial general intelligence (AGI). Experts opine that we remain as far away from AGI as we were years ago, and that making AI chatbots available to the general public does not change this. There have been plenty of anecdotes to illustrate the elusiveness of AGI.

Google’s Bard robbed the company of US$100 billion in market value when it erroneously gave inaccurate answers in a promotional video. Microsoft Bing’s chatbot, which uses ChatGPT, spewed a spate of strange responses that went viral. Companies that are rewarded for their AI developments could just be easily punished by the market when negative news spread. We purport that volatile market swings can occur because our technological progress in AI simply isn’t quite there yet, but the market acts as though we are. This is the third sign of AI speculation.

Before AI, humanity made many advancements that changed the way we live. Agriculture is undoubtably a revolutionising case in point. Even though agricultural practices have been around for over twelve millennia, regulatory bodies like the FDA was only founded in the past two centuries. Even under the supervision of a formal organisation, mistakes still happen. The everchanging state of agriculture with newer innovations has resulted in the FDA bearing the onslaught of its regulatory oversights.

The AI industry will likely face a similar set of issues. Because of its nascency, the support infrastructure is virtually non-existent. Alongside the possibilities that sentient AGIs can unlock, there is much talk about the risks involved. Chief among them is confabulation and misinformation propagation. ChatGPT is driven by plausibility instead of the truth. The chatbot is trained to construct possible output based on a hive of information in its network but has no concept of falsity. Moral dilemmas aside, the potential damage that can be unleashed upon humanity is staggering. The magnitude of growth for AIs that investors are anticipating can only be achieved in parallel with sufficient regulations. However, the system of checks and balances to mitigate the pitfalls of AI chatbots is currently absent. This impedes the rate at which AI companies can expand, which is dichotomous from the unreasonably high valuations we recently witness.

The fifth perspective

There are nonetheless alternatives for investors who still wish to maintain exposure to this burgeoning space. One way is to consider AI exchange traded funds (ETFs), such as the ROBO Global Robotics and Automation Index ETF (NYSE: ROBO) or the Global X Robotics & Artificial Intelligence ETF (NASDAQ: BOTZ). Apart from the usual ETF metrics to look at such as the expense ratio, dividend yield, dividend frequency, and past performance, investors should also pay attention to other points such as the constituents of each ETF.

On top of AI stocks, these ETFs also hold assets in related areas like robotics, cloud computing, and semiconductors, due to the sparse number of pureplay AI stocks. This is not necessarily a bad thing. One of the signs of excessive AI share price buoyancy is the lack of profitability. By including highly profitable behemoths like Nvidia (NASDAQ: NVDA) which are entities that utilise AI, it helps to defray risks associated with companies that are loss-making. Benefits from diversifying across sub-sectors can further be reaped, while still adhering to the theme of investing in AI-related use cases.

Another noteworthy consideration is the geographic dispersion of the ETFs’ holdings. As AI still resides in its embryonic stage, government policies can swing the profitability of a business either way. For instance, businesses which win contracts with government bodies that receive federal funding may be direct or indirect beneficiaries which get to enjoy subsidies and grants. This boosts their chances of achieving breakeven and turning profitable over a shorter time horizon. On the flipside, the wide-ranging capabilities of AIs have security ramifications. As a consequence, geopolitical factors come into play. Depending on the jurisdictions in which these stocks are domiciled and listed in, some of them may be included on certain countries’ blacklist due to sensitivities. It is therefore intuitive to prefer ETFs which consist of more stocks from the former group as opposed to the latter.

There is a narrative that paints a story where humanity’s future is intricately linked to AI. Investors who are strongly compelled to participate in this growth ought to maintain a healthy dose of scepticism. While the scales are tipped in favour of today’s AI leaders, the rankings can be easily scrambled in time to come. The key lies in uncovering and subsequently mitigating the risks. A contrarian approach where investors consistently question their own assumptions and arguments may just be the parachute they need during free fall, after the bubble pops.

Tan Ke Xuan

Ke Xuan holds a Bachelor of Business Management from SMU. He identifies as a value investor who prefers to combine both macro and micro analyses when learning about businesses. He believes there are opportunities to be uncovered in every stage of the economic cycle.

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button