Displacement Risk and the Magnificent Seven

Sammanfattning

This thesis explores the link between technological displacement risk, artificial intelligence-related innovation, and the returns of the largest, high-tech, American firms, the "Magnificent Seven" ("M7"). We suggest that AI innovation is inherently disruptive, gives rise to systematic technological displacement risk, and that its resulting effect on the cross-section of returns can be proxied using the residualized returns of the M7. Theory tells us that this effect varies across value and growth stocks, negatively affecting value stocks, while growth stocks partly hedge the risk associated with innovation activity, resulting in a lower realized value-growth premium as this activity increases. We test the effectiveness of our M7 displacement risk proxy using Fama-MacBeth-style cross-sectional regressions, evaluate its ability to estimate the value premium, and test its correlation with other value factors, as well as "AI" term searches. We find that the factor is not priced (using a t-statistic threshold of 3) in the cross-section when controlling for Fama-French 3-factors, but has the expected effect on value and growth stocks, improves value premium estimates and correlates with the HML and RMW factors, as well as "AI" term searches. The factor also fails to be priced in the cross-section when alternative portfolio sorts are tried, and when momentum, or the CMA and RMW factors are introduced to the model.