In a recent client note, the bank’s strategy team stressed that, while AI adoption is still in its early stages of development, its investment potential remains substantial.
Investors should focus on vertically-integrated players along the AI value chain in the early stages of the AI era. They also highlighted businesses that have a clear monetization path and a strong competitive position.
The AI market has a huge potential. Estimates range from McKinsey $4.4 trillion to Bloomberg $1.3 trillion. UBS estimates that AI-related revenues will exceed $1 trillion annually over the next decade.
AI tools will be expected to drive this growth, as they will improve productivity for knowledge workers who currently number 1 billion worldwide. AI tools such as GitHub Copilot, for example, can help developers code up to 55% quicker, and customer service operations can become 30-50% more effective with generative AI.
UBS has outlined an investment framework that includes three layers of the AI Value Chain: enabling, Intelligence, and Application layers.
The enabling layer includes the physical infrastructure, like AI data centers, that is required for training and running generative AI algorithms. UBS estimates that annual capital expenditures in this layer will reach $331 billion between 2027 and 2028, driven by investments made in AI servers.
UBS notes that AI servers are likely to capture the majority of the value from the enabling layer.
Due to the size of AI computing, the majority of companies will likely consume cloud services. We expect to generate $185 billion of value creation by 2027.
The intelligence layer consists of generative AI algorithms, large language models (LLMs), and computing resources that are derived from the enabling layers. This layer, which is still in its early stages of monetization but plays a foundational role in AI, will continue to grow despite being in its early stages.
UBS stated that “we expect this layer to grow the most into 2027, given its small base.”
UBS strategists have said that the most monetization potential is in the application layer. This includes software applications and services powered by AI. The opportunity is hard to quantify, however, at this stage.
This layer includes tools such as AI co-pilots and personal assistants that have already shown significant productivity gains. Microsoft’s GitHub co-pilot, for instance, generated $100 million in revenue by 2023, and grew by 40% year-overyear with 1.3 millions users.
Strategists wrote: “With developer productivity increases of 50-60%, it is expected that the creation of software code will accelerate.”
UBS believes that the enabling layer is where the biggest opportunities lie in the near-term. The bank continues to expect that the ratio of application layers to the enabling layer and intelligence layer will result in limited bottom-line profits for the application layers during the initial stages.