Which AI wave is worth riding?
By Shawn Lee, Portfolio Manager and Phillip Li, Assistant Portfolio Manager, SGH Australian Small Companies Fund
Welcome to Our US Market x ASX Small Caps Series
As part of our ongoing commitment to delivering valuable insights to Australian investors, Shawn Lee, Portfolio Manager of the SGH Australian Small Companies Fund, recently embarked on an extensive research trip across the United States, meeting with over 40 companies spanning critical sectors like technology, automotive, insurance, and healthcare. This series unpacks the essential takeaways from those conversations, providing a closer look at how American businesses navigate challenges such as inflation, labour constraints, regulatory risks, and the fast-approaching U.S. presidential election. In each article, we explore a specific sector, connecting the dots to help you better understand the opportunities and risks on the horizon.
Stay tuned as we explore each sector, from stocks making waves to advancements in AI and more.
__________________________
Proponents of artificial intelligence (AI) often tout productivity benefits that span virtually all industries, and that will revolutionise how we define work while empowering employees through virtualisation and automation. Generative AI does have significant potential, but only time will tell how much of that potential will be realised.
Like many others, we see the exciting side of developments in artificial intelligence and have positioned our portfolios to take advantage of companies that are early beneficiaries of the AI wave, where we see tangible earning outcomes.
Consideration must also be given to the potential risks arising from AI, such as regulatory risk, consumer backlash, security, and privacy issues.
Opponents of AI conjure up scenarios of existential threats driven by an AI takeover or draw parallels of the current AI boom to previous stock market bubbles and bursts. These aspects of AI are yet to be fully understood. We have yet to fully explore and resolve the many issues and opportunities AI presents us.
As a result, we are cautious about how broadly AI apps can and will be monetised. We believe it will take longer to play out than anticipated as the technology is developed and its utility is determined.
The early beneficiaries of AI
AI training
Large language models (LLMs) are algorithms that can summarise, predict and generate responses in human-like text formats, making life easier for many of us. But before they can be applied to solve a problem or complete a task and raise productivity tangibly, these models need to undergo what is known as ‘training’. This training refers to the initial phase of an AI model where data is fed into a machine learning model and, through trial and error, emerges with a neural network that can consistently solve any challenges presented by the training algorithm.
Once a model is rigorously tested, it can then be deployed for various applications. This phase is called ‘inferencing’; when presented with problems, the AI tool will infer solutions with a relatively high accuracy. This is why we continue to see rapid updates of LLMs, as more training data is processed to enhance their inferencing capability and the overall accuracy of AI tools.
Cloud service providers
Large cloud service providers such as Microsoft (NASDAQ: MSFT), Amazon (NASDAQ: AMZN), and Alphabet’s Google (NASDAQ: GOOG) have been early beneficiaries of AI’s growth. For example, Microsoft recently reported a 24% growth in cloud service revenues, and a meaningful 7% of this growth was attributable to identifiable AI-related demand.
Microsoft is not only shaping up to be an early AI beneficiary but also one that offers significant growth potential. This includes assembling a powerful AI ecosystem outside of its cloud hosting capabilities, deploying AI Co-pilot across various products, investing in data services, leveraging its dominant Office 365 software suite, and its highly strategic stake in the ChatGPT creator, OpenAI.
As we assess the implications of these developments, we anticipate that businesses such as Data#3 (ASX: DTL), one of the largest Microsoft licensing partners in Asia Pacific, will also enjoy these ecosystem benefits over the longer term. However, we remain on the sidelines given the recent management changes and weakening domestic conditions for its enterprise IT sales.
Data centres
Another beneficiary we already see is data centre operators and developers exposed to these large cloud service providers, otherwise known as hyperscalers.
Hyperscalers themselves are pivoting away from a build, own, and operate model when it comes to their infrastructure requirements and are instead leaning more on external data centre operators. This is simply because the time to bring new capacity to market has increased significantly due to lengthening permitting processes and challenges with securing power supply.
In Australia, Infratil (ASX: IFT), which owns a stake in Canberra Data Centre (CDC), is one of the largest data centre developers and operators in Australasia. We expect businesses like CDC to continue to benefit from the theme of hyperscale demand growth over the medium term. NextDC (ASX: NXT) is a similar beneficiary.
Will we see the monetisation of AI in software?
Software companies are another beneficiary of AI; however, there are varying views on whether they can monetise AI apps and features over time.
One example is the human resources and finance cloud platform Workday (NASDAQ: WDAY). It believes all customers should be allowed equal access to AI functionality on its platform. It offers all users at least 50 different AI enhancements without incrementally charging them.
The company’s newly appointed CEO believes some AI functionality will become commoditised over time. In his prior role as a general partner at venture capital firm Sequoia Capital, he observed that many AI startups were building business models based on open-sourced LLMs. While there is an element of being first to market, this competitive advantage gets eroded quickly, given the open-sourced nature of the underlying models. He anticipates that, over time, many of these AI solutions will be housed or replicated by the very customers of these AI startups today.
This view has been echoed by others in the tech industry, including the Australian graphic design platform Canva.
AI adoption and the investment case
There is still a lot to play out in widespread AI adoption. For instance, Microsoft stated that 65% of the Fortune 500 already use their Co-pilot in some form, but only some have rolled this out widely through their organisations.
Other bottlenecks to wider enterprise AI adoption include perennial concerns around data security and privacy and the lack of AI readiness across many organisations. Both software and hardware need to be upgraded to take advantage of AI capabilities.
We are excited about AI developments and have positioned our portfolios to take advantage of companies that are early beneficiaries of the AI training wave, including cloud service providers and data centres, and where we believe there are tangible earning outcomes.
But when it comes to AI deployment more widely, we are cautious and believe this will take longer to play out. We continue to look for more evidence from companies around the monetisation of AI solutions before factoring these into our investment cases.
__________________________
Thank you for reading this article in our US Market x ASX Small Caps Series.
If you’re interested in exploring more, be sure to check out our previous articles, including ASX small caps conquering the U.S. market and our US Market Outlook. Stay tuned for our next article in the series, covering the Automotive industry.
Watch the full video
SG Hiscock & Company has prepared this article for general information purposes only. It does not contain investment recommendations nor provide investment advice. Neither SG Hiscock & Company nor its related entities, directors or officers guarantees the performance of, or the repayment of capital or income invested in the Funds. Past performance is not necessarily indicative of future performance. Professional investment advice can help you determine your tolerance to risk as well as your need to attain a particular return on your investment. We strongly encourage you to obtain detailed professional advice and to read the relevant Product Disclosure Statement and Target Market Determination, if appropriate, in full before making an investment decision.
SG Hiscock & Company publishes information on this platform that to the best of its knowledge is current at the time and is not liable for any direct or indirect losses attributable to omissions for the website, information being out of date, inaccurate, incomplete or deficient in any other way. Investors and their advisers should make their own enquiries before making investment decisions.
