If 2023 was the year for AI excitement, this year may be the year for deployment. In first quarter earnings calls, approximately 45% of S&P 500 companies mentioned AI, marking a fresh high by our measures, and their collective investments continue to climb. S&P 500 capital expenditures are set to accelerate to a 7% y/y pace in Q1, up from 4% in Q4, while consensus expects the five major AI hyperscalers^[1] to spend nearly $200 billion on capex this year, representing a 37% y/y increase and over 10x growth in the last decade. As the investment “railway tracks” are being laid, business leaders across industries are solving for the necessary infrastructure, skilled labor and workflow realignment needed to adopt AI technologies.

The financial services industry is no exception to this trend. Just as the emergence of fintech enabled enhanced algorithmic trading and customized portfolio management, artificial intelligence has the potential to drive a new wave of opportunities to improve efficiency, customization and agility in investment management.

Consider three key applications:

• Enhancing the research process. A defining feature of AI is its ability to synthesize vast amounts of data. Apply this skill to decades of a financial firm’s proprietary data and over 100 terabytes of public data, and an AI large-language model (LLM) can help research analysts separate the signal from the noise, improve forecasting models and augment their fundamental process with sophisticated quantitative and textual analysis.
• Optimizing portfolio management. AI tools can “coach” investment professionals by analyzing their historical investment decisions and providing personalized and actionable insights, saving valuable time and mitigating biases. AI may also help optimize asset allocation by improving risk and return estimates and fine-tuning portfolios to highly customized targets and constraints.^[2]
• Making trading more efficient and informed. Given the increasing size, frequency and complexity of trades, AI can help traders reduce transaction costs and improve execution. For instance, an LLM interface can quickly provide traders with relevant pricing data across exchanges while helping identify the best time, size and venue for trades.

While exploring these business applications, financial firms must also address the associated risks, such as data security, data quality and ensuring appropriate human supervision to identify errors and properly contextualize AI results. Adoption will require some trial-and-error, worker training and improved access to tailored applications. However, much of this business discovery appears underway. For instance, at J.P. Morgan, technology spend is expected to reach $17 billion this year^[3], a 10% increase from the $15.5 billion spent in 2023, and a taskforce of AI and machine learning professionals are actively solving for use cases across business verticals.

Ultimately, as markets expand their focus beyond AI infrastructure providers to broader AI adopters, investors should take care to separate AI buzz from successful and profitable adoption. Beyond investment management, opportunities for adoption will be widespread and the “railway tracks” being built will require resources and expertise from a vast array of companies and sectors in the economy. 

^{[1] Hyperscalers, also known as hyper-scale cloud providers, refers to largest tech companies that operate massive data centers and provide scalable cloud computing services required to run Generate AI tools. The major 5 companies referenced are Amazon (AWS), Microsoft (Azure), Meta, Oracle and Alphabet (Google Cloud).}

^{[2] See for instance, Sohnke M Bartram, Jurgen Branke and Mehrshad Motahari, “Artificial intelligence in asset management”, CFA Research Institute, 2020.}

^{[3] JPMC, 2024 Investor Day.}

^{09pz242105202006}