Should investors worry about market concentration?

Market Concentration Metrics and Historical Context

• Current concentration levels: The top 10 stocks by market cap currently comprise approximately 36% of the S&P 500, a figure significantly higher than the 20% long-term average and 25% peak during the 2000 dot-com boom.
• Historical comparison: Using data dating back to 1932, one specific concentration metric currently reaches levels unseen in nearly a century.
• International and historical diversification: Owen Lamont notes the U.S. market is more diversified than current European markets and significantly less concentrated than historical markets (e.g., 1930s-1950s) or foreign markets like Taiwan and South Korea.
• Driver of concentration: Lamont attributes current concentration primarily to the mechanical byproduct of profit concentration, where the top 10 firms have generated an outsized share of total profits over the last decade.

David Kostin's Bearish Case on Concentration and Returns

• Return forecast shift: Kostin projects 10-year forward annualized returns to range between -1% and 7%, with a midpoint of 3%, compared to a baseline historical average of 11% and a recent 10-year average of 13.5%.
• Impact of concentration variable: Incorporating market concentration as a variable in the forecasting model lowers the projected return midpoint from 7% to 3% (a 400 basis point reduction) by statistically signaling lower forward returns.
• Valuation disconnect: The top 10 stocks trade at 31 times earnings (an earnings yield of ~3.2%), which is below the 10-year U.S. Treasury yield (~4.2%), resulting in a negative equity risk premium not seen in 20 years.
• Growth sustainability risk: Kostin argues that while these firms are expected to sustain ~20% annual growth, historical data shows almost no companies can maintain such growth rates over a decade; the probability of this occurring fades dramatically over time.
• Volatility warning: High concentration implies higher realized forward volatility due to reliance on a narrow group of companies, yet investors are not compensated for this risk as valuations are not attractive.
• Primary investment recommendation: Non-taxable investors should favor equal-weighted benchmarks over cap-weighted indices for 10-year horizons, as equal-weighted indices have outperformed cap-weighted indices 80% of the time over such periods in the model.

Owen Lamont's Counter-Argument on Risk and Returns

• Concentration vs. Risk: Lamont argues that market concentration does not inherently increase market risk; he distinguishes between a concentrated portfolio (higher risk) and a concentrated market (which may be safer due to economic fundamentals).
• Valuation over structure: Lamont posits that poor future returns for large stocks are driven by high valuations and subsequent mean reversion (the "value effect"), not the concentration itself.
• AI and market disruption: Lamont highlights AI as a double-edged sword with high upside and downside potential, capable of either creating a bubble similar to 1999 or destroying value for existing firms through "creative destruction."
• Future market composition: Lamont predicts that today's dominant firms will not be the primary value drivers in 20 years, as the "magnificent seven" are unlikely to maintain dominance indefinitely due to market turnover (roughly 33% index turnover over a decade).
• Geopolitical and macro risks: Lamont identifies geopolitical tensions and the transformative impact of AI as significant risks that could drive significant market volatility, independent of concentration metrics.

Converging Forecasts and Divergent Conclusions

• Agreement on lower returns: Both strategists agree that the U.S. equity market is currently overvalued and likely to deliver lower returns over the next decade compared to the previous 10 years, though they attribute this to different causes (concentration for Kostin; general valuation for Lamont).
• Divergence on actionable advice: Kostin advocates for a strategic shift to equal-weighted portfolios to mitigate concentration risk, whereas Lamont suggests the risk lies in the valuation of growth stocks and general market cycles rather than the weight distribution.
• Forward-looking variables: Both analysts cite the sustainability of current growth rates and the potential for a bubble in the AI sector as key unknowns that could significantly alter 10-year return profiles.
• Risk factors identified: Kostin cites potential household allocation increases and AI-driven sales sustainability as risks to his bearish thesis, while Lamont cites the possibility of an AI boom or bust as the primary source of future variance.