David Hay of Haymarket Capital/David Hay Research has highlighted this counter-intuitive but well-documented phenomenon. Key points from the data primarily based on research covering ~1990–2022, with similar patterns persisting in many later studies.
Stocks that are deleted from the S&P 500 the “black line” in his chart — typically because they no longer meet size, liquidity, or profitability criteria — have historically outperformed the S&P 500 index itself by roughly +3% to +5% per year on average after deletion.
Over the 32-year period he references 1990–2022, this translated into ~400–500% cumulative outperformance versus just staying in the index. Forced selling distortion: When a stock is removed from the S&P 500, index funds and ETFs hundreds of billions of dollars must sell it immediately, regardless of fundamentals.
Register for Tekedia Mini-MBA edition 19 (Feb 9 – May 2, 2026): big discounts for early bird.
Tekedia AI in Business Masterclass opens registrations.
Join Tekedia Capital Syndicate and co-invest in great global startups.
Register for Tekedia AI Lab: From Technical Design to Deployment (next edition begins Jan 24 2026).
This creates temporary downward pressure and often leaves the stock undervalued. Mean reversion / value effect: Many deleted stocks are “fallen angels” — former large-cap growth darlings that have underperformed and become relatively cheap low P/E, high dividend yield, etc.. Value and quality factors tend to rebound over time.
After deletion, the average market cap drops dramatically often from >$10–20 bn to <$5 bn. Smaller stocks have historically outperformed larger ones over long periods. The S&P 500 is constantly pruning its weakest members and adding new winners. So by construction it looks better than it actually is for a buy-and-hold investor in the “average” constituent.
The deleted stocks are the ones that got kicked out, yet ironically many recover. Real-world performance approximate, from various studies: 1-year average excess return after deletion: +6–12%. 3-year average excess return: +3–5% annualized.
Long-term (5–10 years): still positive but narrower. A mechanical strategy of buying every S&P 500 deletion equally or float-weighted and holding for 1–3 years has beaten the S&P 500 by 300–600 bps annualized in most backtests since the 1980s, with higher volatility and drawdowns.
Not every deleted stock recovers some go to zero — think Sears, Toys“R”Us, etc. Transaction costs, liquidity risk, and tax implications reduce real-world returns. The edge has narrowed in recent years as more investors quant funds, factor ETFs have started exploiting it.
But the core observation remains valid and is one of the cleaner examples of index mechanics creating exploitable inefficiencies. David Hay is correct: over the long run, the “reject pile” of the S&P 500 has actually been a better place to fish than the index itself.
The Russell 2000 has shown signs of resurgence, with year-to-date (YTD) gains of approximately 9.85–10.45% (based on ETF trackers like IWM and VRTIX), outpacing its historical average but trailing the S&P 500’s stronger tech-led rally.
Over the past six months, it has gained ~17%, matching the S&P 500 more closely, amid expectations of Federal Reserve rate cuts and a shift toward “animal spirits” in the market.
The Russell 2000’s returns are more volatile than the S&P 500’s, with larger drawdowns but potential for outsized gains during rotations. Since 1979, the two indices have had a high correlation ~0.8 on average, but divergences highlight small-cap cycles.
Small caps have outperformed large caps about two-thirds of the time since 1927, per long-term data, though the edge has been negative since 2019 due to the “Magnificent 7” tech dominance.
Russell 2000 beating S&P 500 by notable margins: 1979–1983: +77–80% relative gain amid double-dip recessions, high inflation, and early 1980s recovery. 1990–1994: +49.6% during the 1990–1991 recession and early expansion.
1999–2006: +99% through the dot-com bust 2000–2002 recession and initial 2003–2007 expansion. July 2024: +10 percentage points best monthly relative gain since Feb. 2000, driven by rate cut bets.
Post-2024 Election Cycles: Historically +6–12% relative in the 6–12 months following U.S. elections, due to pro-domestic policies. 1983–1990: -91.4% relative S&P dominated in 1980s expansion with high real rates.
1994–1999: -94.5% amid late-1990s boom and rising rates. 2019–2024: Consistent lag (e.g., -6.8% over 3 years ending 2024), as S&P benefited from ~30% tech weighting vs. Russell’s ~4%.
The S&P SmallCap 600 a profitability-filtered small-cap index has outperformed the Russell 2000 in 14 of 21 years since 1994, highlighting how the Russell includes more unprofitable “junk” stocks that drag returns. Small-cap rallies like the current one are often short-lived but explosive.
~90% of Russell 2000 revenues are domestic, making it a pure U.S. economy bet. It thrives in expansions or post-recession rebounds when risk appetite rises, as smaller firms are nimbler and benefit from catch-up growth.
Small caps are highly leveraged ~50% of debt is short-term/floating rate, so Fed easing reduces refinancing costs more than for large caps. Historically, the first rate cut in a cycle boosts small caps by 5–10% relative to large caps. With markets pricing ~90% odds of a September 2025 cut, this supports further gains.
Russell 2000 trades at a discount (e.g., P/B in bottom decile since 1990; higher dividend yields). After lagging (e.g., due to 671/2,000 unprofitable firms vs. 25/500 in S&P), rotations into “cheap cyclicals” financials, industrials, healthcare—~48% of index drive rebounds.
Less tech exposure ~15–17% vs. S&P’s 30% means outperformance when cyclicals lead like the post-vaccine yield rises in 2020–2021. Value and equal-weight tilts amplify this. +11% in five days in July 2024) reflect “greed” and broad participation, but overbought signals (RSI >79) often lead to pullbacks.
The Russell 2000 hit 2,500 for the first time in October 2025, up ~17% in six months, but trails the S&P 500 YTD due to AI/tech persistence. Bank of America sees continued outperformance through September 2025 absent tariff shocks, tied to earnings recovery Q2 profits beat expectations, ending a “profits recession.
However, October seasonality is weak for small caps, and speculation risks a correction. Compared to the S&P SmallCap 600 +11% in six months, the Russell’s 9% edge signals elevated “animal spirits” but potential froth. If U.S. growth outpaces global peers and rates fall, small caps could extend gains historical post-election edge + recovery regime.