Beyond the AI Selloff

To start the year, investors were already navigating a volatile market environment. Since November, cyclicals, value, and the equal weight version of the S&P 500 Index had been outperforming cap weighted benchmarks and growth stocks, signaling a broadening in market leadership after a narrow mega cap driven rally. We noticed that this shift prompted significant repositioning and drove elevated volatility across single stocks, factors, and themes. Into this unsettled backdrop, rotational forces were exacerbated by rapid advances in generative AI, large language models, and autonomous workflow tools that have added a new source of disruption.

What has changed most, in our view, is the narrative. AI is no longer being treated simply as a productivity enhancer. Increasingly, it is being priced as a direct competitor, one that could pressure revenue models and margins across established industries. From 2023 through 2025, markets largely rewarded incumbents on the assumption that AI would reinforce existing business models. In early 2026, that assumption shifted abruptly. Entire sectors began trading as though displacement were certain and immediate. We think many of these moves reflect crowded positioning and short-term fear more than realistic adoption timelines. Still, these episodes are important reminders of how quickly sentiment can reset and why risk management is critical when narratives move faster than fundamentals.

Case Study 1 | Disruption in the software subsector

After the launch of new AI workflow tools from Anthropic and initiatives at Alphabet Inc., investors assumed that AI agents could quickly replace enterprise software subscriptions.

• The S&P North American Technology Software Index fell 15% in January, its largest monthly decline since 2008.

• Since 2020, this index weight made up of growth compounders, many of which were SaaS companies, has increased materially, and they became a large part of many investors’ portfolio

We continue to watch this software subsector for opportunities as indiscriminate selling has been driven more by sentiment than fundamentals. In many cases current valuations reflect disruption that may unlikely unfold at the speed implied by prices.

Case Study 2 | Disruption in Data, Analytics and Business Services

Investors reasoned that if AI can automate legal and analytics workflows, then billing, coding, and health information businesses are also at risk. Selling became broad and indiscriminate. However, our work shows that real world change is often slow due to regulation, integration challenges, and liability considerations. Many of these companies operate within complex systems that cannot be replaced overnight, even if the technology exists.

Case Study 3 | Disruption in the Insurance and Brokerage sub sector.

We have also seen that new AI powered tools raised concerns about disintermediation in the insurance and brokerage services space.

• Insurify launched an AI shopping agent that scans millions of quotes, checks credit, reviews records, and prepares contracts.

• At the same time, Spanish digital insurer Tuio introduced the first AI insurance app approved on the ChatGPT store

• Altruist rolled out Hazel AI for wealth management and brokerage

The S&P 500 Insurance Index recorded its largest drop since October, at -3.9%. Yet when we step back, underwriting discipline, capital management, and regulatory frameworks remain the core economic drivers of the industry. Distribution models may evolve, but the economics of risk, pricing, reserving, and capital allocation, do not disappear simply because the interface changes.

We expect these rolling waves of disruption and dislocation to continue. The market is still trying to determine which companies will use AI as a tool and which may ultimately be displaced by it.

In this type of environment, our process becomes even more important. We aim to combine bottom-up fundamental work with quantitative analysis, always anchored in a risk-first framework. We spend significant time separating what is technically possible from what is economically probable by breaking down revenue exposure in detail, testing assumptions about adoption against actual customer incentives, and incorporating positioning data to identify when forced selling distorts prices. In periods of rapid narrative shifts, that discipline helps us find opportunity where others may see only threat.

When narratives shift quickly, discipline and patience matter. Our job is not to react to every headline, but to assess how much of the future is already reflected in the price, and where risk and opportunity may have become misaligned.