The views expressed are those of Anchor Capital Advisors, LLC (“Anchor”) and are subject to change at any time. They are based on our proprietary research and general knowledge of said topic. The below content and applicable data are in support of our views on said topic. Please see additional disclosures at the end of this publication.
Introduction
Artificial Intelligence (AI) is poised to be one of the most transformative and disruptive technologies of the next decade, with far-reaching implications across various industries and sectors. Its economic potential has already helped lift the stock market to new highs.[1] However, as with anything accompanied with this much hype, it introduces risks to the market and investor’s portfolios that can often get overlooked. This piece explores the significance of AI as a key investment theme for the coming years, its potential risks and rewards, and how Anchor’s investment team is approaching the upcoming technological revolution.
The Rise of AI
Artificial Intelligence refers to the ability of machines and computers to perform tasks that would normally require human intelligence, such as learning, reasoning, identifying patterns, problem-solving and making decisions.[2] For many of us, AI is already integrated into our daily lives without even realizing it. It’s behind the facial recognition technology that unlocks an iPhone, new song recommendations on Spotify, chatbots that provide customer service, and the fastest route suggestions on Google Maps.[3] In fact, the history of AI dates back to the 1950s, so you might be wondering, why is everyone so hyper-focused on it now? The short answer: ChatGPT.
If we rewind back to 2022, the Covid stimulus bubble appeared to be bursting. Throughout 2021, the market had been displaying many of the classic signs of a bubble peaking including extreme investor euphoria and a flurry of companies without viable business models rushing to IPO or SPAC. So-called “meme stocks,” such as GameStop and AMC, surged from a dramatic influx of buying activity by retail investors. Ultimately, widespread inflation and aggressive interest rate hikes sparked recession fears that led to profit taking and enough consternation to catalyze a stock market drawdown of more than 25% in the first ten months of 2022. In fact, the first half of 2022 saw the S&P decline more than in any first half of a year since 1939.[4]
However, just as calls for a recession intensified and we officially entered a bear market in late 2022, it was interrupted by the launch of ChatGPT and with it, the promise of AI. ChatGPT is a large language model (LLM), a form of so-called generative AI, that can find, synthesize, and contextualize information it finds on the web to provide answers. Generative AI refers to AI models that can create entirely new content like text, images, audio, and code from scratch, rather than just analyzing existing data as was the case with previous forms of AI. ChatGPT ended up becoming the fastest growing consumer application of all time with over 1 million users in the first five days and 230 million users in the first six months, capturing public imagination and widespread media attention for its ability to engage in natural conversation, provide detailed explanations, and assist with a wide range of tasks.[5] Once OpenAI, the company behind ChatGPT, demonstrated the potential of generative AI and LLMs and the world saw the magnitude of the consumer response, it sparked a surge of investor excitement and funding for AI. This
helped lift the S&P out of its bear market and ultimately to new all time highs in late 2023 that have extended to date in 2024.
Exhibit 1: Unprecedented Speed to Adoption [6]
Source: https://a16z.com/the-economic-case-forgenerative-ai-andfoundation-models/
The AI Infrastructure Buildout and Its Investment Implications
Investing behind a new and powerful theme, especially one with as much hype and attention as AI, can be tricky. We have studied history and found that almost every new technological paradigm shift – like the one we are seeing today with AI – typically follows a similar pattern going back from the internet to telephones to railroads.
Phase 1: Infrastructure Buildout
The first stage is the infrastructure buildout phase, which in the case of AI involves setting up the physical environment, such as data centers, that will empower organizations to effectively develop, deploy, and manage AI projects. This infrastructure is being built primarily by the hyperscalers – microsoft, Amazon, Meta, and Google – and will most likely serve as the backbone for future AI platforms. To put into context just how much money these companies are spending – in 2024, the combined CapEx from just those four companies alone is expected to exceed the other 90 technology and communication companies within the S&P 500.[7] The buildout requires copious amounts of the latest chips, servers, networking and storage equipment. Thus, during this phase, it is the companies
that sell the hardware and materials that benefit. As the saying goes, when everyone digs for gold, sell shovels. Nvidia, with its AI chips, has undoubtedly been the poster child for this first stage.
Investment Implications
However, what ends up typically happening in this stage is that investors start extrapolating the rapid growth rates being experienced by these companies in the short-term, far into the future. The companies become overvalued but eventually the infrastructure buildout slows, or macro conditions change, causing the growth rates of these companies to decelerate. The market then realizes the lofty expectations about the future that are built into those high valuations won’t be realized, and the stocks drawdown viciously to reflect that.
Example:
Semiconductor stocks are known for their violent cycles. As Todd Ahlsten at Parnassus Investments explains: “When product or technology innovations happen, they create rising demand for the new chips that will be needed to power those innovations. The more chip demand exceeds available supply, the more likely customers are to place double or triple orders, pushing prices higher and potentially creating a false sense of demand. Once the capital investments have been completed and the manufacturing facility is up and running, chipmakers can see revenue, profits and stock prices soar. Often, they will over-extrapolate by adding excess capacity to meet demand. As supply meets and exceeds demand, customers will typically stop placing orders or cancel them. Chipmakers, saddled with excess inventory, slow or stop manufacturing until orders pick up again. During this part of the cycle, revenue, earnings and stock prices typically fall significantly.” The industry is in “a constant rotation between chip scarcity and oversupply, which, in supply chain terms is called ‘the bullwhip effect.’”[8] We don’t have to look back far to see how this bullwhip effect can impact the prices of semiconductor stocks. Semiconductor stocks experienced a strong upcycle coming out of Covid until early 2022 but then the cycle slowed and the semiconductor industry, as measured by the iShares Semiconductor ETF (SOXX), fell by nearly 50%. Nvidia and Advanced Micro Devices (AMD) both fell over 65% from peak-to-trough during this time.[9]
As evidenced above, investing in high growth stocks is inherently risky and volatile. This is due in part to their high valuations, which can change rapidly and without warning, resulting in major drawdowns. To invest in these companies, you must make sure you are prepared to suffer through the large drawdowns that can take place. Unfortunately, studies have shown time and time again that most people aren’t prepared for this and react poorly to volatility.[10] They end up buying and selling at the most inopportune times and their investment returns suffer as a result.[11] It’s easy to say you will have the mental fortitude to hold onto a stock during a 65% drawdown because you believe strongly in its future, but it’s much harder to do than you’d think. The reality is that major drawdowns are extremely painful and the only way to stop that pain in the short-term is to sell. Once you sell, it is very hard to get back in and often results in missing a significant portion of the upside recovery, if not all of it.
Off the October 2022 lows, the semiconductor index has already rallied over 125% and the broader technology index, as measured by the Nasdaq Composite Index, has rallied over 80%. Nvidia is up more than 10-fold. Many of these companies are once again trading at nosebleed valuation levels, relative to their normalized earnings power. That’s normalized earnings power, not peak earnings power – a key distinction in assessing a company’s valuation. Thus, investing in these stocks today is a much different decision than it was 20 months ago. The AI infrastructure buildout and related upcycle could keep going on for longer but the risk-reward after such a strong run, knowing the history the history of major drawdowns for these stocks, does not seem favorable in our opinion. We find it amazing how quickly the market has forgotten the perils of growth investing and the lessons of 2022.[12]
Phase 2: Creating New Products and Services
The second phase of AI will come from companies that create new products and services using this new hardware and infrastructure being built in phase one, just as mobile apps flourished after smartphones and wireless communications became ubiquitous. This is where the real, long-term value creation will take place. Where completely new business models will be invented. Where the next Facebook or Google or trillion-dollar company could emerge. In 10-20 years, it is possible that we look back and realize that AI was every bit as powerful and world changing as the internet, if not more so due to the innovation and productivity it may bring.
Investment Implications
The problem that arises in this phase is that, in the short-term, it’s likely that these AI applications will be more hype than reality. Many of the products and services may be theorized but don’t exist today. Unlike Nvidia who is seeing real, tangible benefits from AI today, the companies in this stage either don’t exist yet or are not seeing any additional revenue or earnings from AI currently. However, investors have imagined the ultimate possibilities and already begun to price much of the very long-term potential of the technology into current market prices.
In our opinion, we are at the point in the AI cycle where all investors see is opportunity and upside and no risk. They are scrambling to get exposure to AI at any cost due to the fear of missing out (FOMO). But as we have seen with past technology revolutions, there is usually a substantial period of disappointment in the markets even though such revolutions often end up being just as transformative as those early investors hypothesized, sometimes even more so. Bill Gates famously said, “Most people overestimate what they can achieve in a year and underestimate what they can achieve in ten years.” In this case, we think there is a good chance the market is overestimating the impact AI will have in the short-term and underestimating the impact it will have over the long-term.
Example:
Perhaps the most famous example of this is in history is Amazon. Amazon’s stock price rose more than 50 times from its IPO in 1997 until its peak in 1999, but then declined by over 90% in the subsequent three years when the dot-com bubble burst.[13] The company went on to create tremendous value for the world and for shareholders over the next 20 years but for those that invested in the late 1990s, it took more than 10 years for their investment just to breakeven. It’s likely many of those investors sold at some point during those 10 years, which means that they lost money despite being right that Amazon would go on to be a transformative company and monopolize nearly half of the retail world. Amazon may be the extreme, but we think the point still holds. Price always matters.
Today, many companies may be using the “AI” buzzword on earnings calls with investors but won’t be a part of the revolution taking place. Others will, but we believe the timelines provided by these management teams to see the benefits of AI will be longer than expected. Most market participants today are impatient and possess short-term time horizons, which leaves these companies’ stock prices vulnerable to even the slightest misstep in their AI journeys.
How Anchor is Approaching AI
At Anchor, our primary focus is on capital preservation. We believe we can deliver a competitive set of returns over a full market cycle but with greater downside protection. Our 40+ years of experience in the investment business informs our view that periods of rapid technological innovation and investor enthusiasm can drive stock prices higher than company fundamentals justify. As such, Anchor’s investment team has been taking a measured approach to investing in AI across our portfolios, keeping in mind the uncertainties created by the unprecedented pace of change taking place, not extrapolating near-term results too far into the future, and remembering that price matters.
- Owning companies directly benefitting from AI infrastructure buildout. We have owned several traditional value stocks that have directly benefitted during this AI infrastructure buildout phase, such as one that sells electrical components to AI data centers, but we have sold or trimmed many of these investments as we believe they have become overvalued and where the downside is now greater than the upside. We also own numerous companies who we believe will benefit tremendously from AI down the road, but with that potential upside not priced into their stocks today. For instance, we own a major agricultural equipment manufacturer that is leveraging AI and machine learning to develop autonomous tractors and other intelligent farming tools to optimize crop yields and reduce labor cost for farmers but is trading at an all-time low valuation. Several utilities and gas producers in our portfolios stand to benefit from the immense energy needs of AI data centers, but trade at valuations far below the market average.
- Owning companies indirectly benefitting (i.e. AI boosting productivity). We have also observed that the market has been mostly focused on the revenue opportunity associated with AI and has been slower to appreciate the potential cost benefits it will bring to companies. Emerging AI products and services could not only bring sales and earnings to the companies that create them, but they could also benefit companies across a range of industries that can use them to boost productivity. We believe value stocks will be huge beneficiaries of the productivity gains that will come from AI given their greater focus on expenses and profits vs. revenue growth. Those productivity gains will be permanent increases to their bottom lines whereas with growth stocks, their newfound AI revenues will need to be replaced every year.
- Using AI to identify winners, but also detractors. Last and certainly not least, while our research team has been actively seeking out companies poised to thrive by leveraging AI, we have also been diligently analyzing potential disruptions and scrutinizing whether AI could impair our existing investments. We have been monitoring for new AI-driven entrants that could shake up certain industries just as internet companies like Amazon did to many brick-and-mortar retailers over the last few decades. This is the side of AI portfolio investment discussions we think is getting most overlooked. Our team aims to not only identify the winners propelled by AI but also avoid the potential losers displaced by this transformative technology.
Conclusion
AI is a powerful theme that will have a profound impact on nearly every industry and sector of the economy. From healthcare to transportation to finance and manufacturing, AI will most likely drive innovation, boost productivity, and create new business models and revenue streams.
For market participants, AI represents a powerful long-term theme to invest behind. Companies that successfully develop and/or deploy AI technologies will, in our opinion, be well-positioned to gain market share, improve margins, and deliver strong returns to shareholders. At the same time, AI and the investor euphoria that surrounds it has elevated the risks in the market and for particular growth stocks. Anchor is working hard to steer clear of these risks while still capturing the upside that AI can bring to our portfolios. We believe this dual focus is critical to capitalizing on AI’s immense potential while preserving our client’s capital in the process.
[1] https://www.troweprice.com/financial-intermediary/us/en/insights/articles/2023/q2/generative-ai-holds-promise-and-peril-for-investors.html
[2] https://www.nvidia.com/en-us/glossary/generative-ai/
[3] https://www.wsj.com/articles/no-business-plan-no-problem-chatgpt-spawns-an-investor-gold-rush-in-ai-6bdbed3c
[4] https://www.gmo.com/americas/research-library/the-great-paradox-of-the-u.s.-market_viewpoints
[5] https://www.fastcompany.com/91092373/big-tech-is-on-a-generative-ai-hiring-spree
[6] https://a16z.com/the-economic-case-for-generative-ai-and-foundation-models/
[7] Bloomberg and Bernstein Analysis
[8] https://www.investmentweek.co.uk/opinion/4166409/parnassus-investments-cio-anatomy-semiconductor-cycle
[9] FactSet financial data and analytics
[10] https://sg.morningstar.com/sg/news/240091/heres-why-you-underperform-the-fund-you-own.aspx
[11] Ibid
[12] FactSet financial data and analytics
[13] FactSet financial data and analytics
The views expressed are those of Anchor Capital Advisors, LLC (”Anchor”) as of the date written and are subject to change at any time. Anchor does not undertake any obligation to update the information contained herein as of any future date, nor does it have liability for decisions based on this information. Certain information (including any forward-looking statements and economic and market information) has been obtained from sources we deem reliable, but is not guaranteed by Anchor, nor is it a complete summary of available data. This publication has been prepared by Anchor Capital Advisors, LLC (Anchor). The information is for educational purposes only and should not be considered investment advice or a recommendation of any particular strategy or investment product. These opinions are not intended to be a forecast of future events or a guarantee of future results. No part of this document may be reproduced in any form, or referred to in any other publication, without express written permission of Anchor. Past performance is not guarantee of future results. Inherent in any investment is the possibility of loss. The benchmark returns include in reinvestment of income. Time-weighted portfolio returns are calculated for each monthly period in the prior quarter. Quarterly results are linked to determine annual returns. Individual client portfolio results may vary from the results presented for the model because of different investment objectives, tax status and other considerations. Returns of individual client accounts will be reduced by advisor fees and other expenses which might be incurred to provide investment management, custody, administrative, actuarial, accounting or other services to the client. A complete list of each security that contributed to performance is available upon request.