AI Fear and Insurance Stocks

Anthropic released a handful of Claude Cowork plugins last Friday that caused a subsequent slide in the price of certain insurance stocks. Part of the insurance-related sell-off is attributed to the potential threat that Claude’s 11 plugins may pose to the brokerage and consulting businesses of insurance companies such as Aon, Marsh & McLennan and Willis Towers Watson.

The fear is that Cowork, an agentic AI assistant, will handle everyday workplace tasks that could obviate, or at least reduce demand for, the risk placement services and the consulting services that insurance brokers provide. To assess the impact of the potential threat, this week’s report points out that revenue from brokerage and consulting businesses is 100% of total revenue for certain insurance brokers affected by last week’s slide. We also point out that other companies, such as P&C companies, may realize efficiency gains from AI advancements, but upside would likely be limited to a segment of total expenses.

Macro Trends

Over the last week, macro data was released that points to an improvement in the jobs environment. If this environment is sustained, it will support the earnings of insurance companies with significant books of group business, such as Met, Pru, and Hartford.

• February 7 initial jobless claims and unemployment: decreased to 277,000, which was slightly above expectations and remains at historically low levels.

• February 11 nonfarm payrolls: added 130,000 jobs, which was above expectations, and unemployment remained at 4.3%.

Combined, the data points support a welcome improvement in the demand for labor. The red box in the figure below highlights the recent multi-month uptick in unemployment that sparked concern for insurance companies with exposure to group business such as Hartford, Met, and Pru. The most recent data, while certainly not a trend, provides welcome respite from months of tepid labor data. The improved jobs data sets the stage for what could be a positive 1Q environment for Hartford, Met, and Pru.

A sustained uptick in the jobs data would help reverse the tailwinds that, as we addressed in a prior note (Insurance: Contrary to Positive Estimates, Jobs Data Points to Headwinds), would impact companies with exposure to group premiums, such as HIG, MET and PRU. The following table links macro data, such as the unemployment rate, to insurance company earnings.

Linking Macro Trends to Potential EPS Impact

Our Macro Tracker table lists key economic data relevant to insurance company earnings. The right-hand column ties macro trends to the potential impact on company earnings.

Deep Sector data and functionality shown in this report are available through the FactSet Workstation.

Downward Pressure on Insurance-Broker Stocks

Anthropic’s release of its Claude Cowork plugins last Friday poses a direct threat to the brokerage and consulting businesses of insurance companies such as Aon, Marsh & McLennan, and Willis Towers Watson. The figures below, sourced from respective press releases, split total revenue between brokerage and consulting businesses.

Given 100% of the insurance broker revenue for the stocks below is derived from two insurance businesses considered under threat from advances in AI, investor concern is not unwarranted if the threat emerges as credible.

Potential Upside for Insurance Stocks?

Anthropic’s release of its Claude Cowork plugins could pose a direct benefit to P&C companies in terms of productivity. As one example, P&C companies could improve claims processing productivity in terms of both time and costs. However, unlike insurance brokers that could see downside to 100% of revenues, and were repriced accordingly, P&C companies could see upside, but it would only impact a portion of the total cost structure.

This blog post is for informational purposes only. The information contained in this blog post is not legal, tax, or investment advice. FactSet does not endorse or recommend any investments and assumes no liability for any consequence relating directly or indirectly to any action or inaction taken based on the information contained in this article.