Week 51 Day 3: Chasing Performance: Last Year's Winner Is Next Year's Loser

The performance-chasing cycle: (1) A fund or asset class has a great year. (2) Financial media features it. (3) Investors pile in. (4) The fund underperforms over the next several years (mean reversion + bloated size). (5) Investors sell at a loss and chase the next hot fund. Repeat forever. Real example: In 1999-2000, technology funds returned 80-100%. Billions poured in. Over the next 3 years, those funds lost 70-80%. The investors who bought at the peak never recovered. Meanwhile, value funds that everyone ignored in 1999 outperformed dramatically from 2000-2006. The hot fund of 2024 will almost certainly underperform the market average over the next decade. You just do not know which hot fund it is yet -- and that is the point.

The empirical evidence against performance chasing is overwhelming. Morningstar assigns star ratings based on past risk-adjusted returns: 5 stars for the top performers, 1 star for the worst. Kinnel (2010) showed that 5-star funds subsequently underperformed 1-star funds over the next 5-10 years. The reason: star ratings measure past returns, and past returns mean-revert. A fund that beat its benchmark by a wide margin often did so through concentrated bets or style drift that are unlikely to persist. Meanwhile, the 1-star funds were often simply out of favor, holding assets that were cheap and poised for recovery. Frazzini and Lamont (2008) documented 'dumb money' flows: investors collectively move money from assets with high expected returns (recently underperforming) to assets with low expected returns (recently outperforming), reducing their aggregate returns by approximately 1% per year. This behavior is driven by three cognitive biases: (a) representativeness -- investors extrapolate recent trends, assuming a fund that returned 30% last year will return 30% next year; (b) availability bias -- recent high performers are featured in media, making them cognitively 'available' as investment choices; (c) regret avoidance -- buying a recent winner feels safe because if it continues to perform, you look smart, and if it fails, 'everyone' was buying it. The antidote: invest in index funds that own everything. You cannot chase performance when you own the entire market. You are automatically invested in next year's winner (and next year's loser), and the winners always outweigh the losers over time.

The persistence (or lack thereof) of mutual fund performance has been one of the most intensively studied topics in empirical finance. Carhart (1997) showed that after controlling for the Fama-French three factors plus momentum, the average mutual fund had an alpha of negative 1.8% per year -- meaning funds systematically destroyed value relative to passive factor exposure. More importantly, Carhart showed that the apparent persistence of top-performing funds was entirely explained by momentum (the tendency of recent winners to continue winning for a few months), not by manager skill. Once the momentum factor was controlled for, persistence vanished completely. Berk and Green (2004) provided the theoretical explanation for why even genuinely skilled managers cannot deliver persistent outperformance to investors: success attracts capital inflows, increasing fund size until diseconomies of scale (market impact, limited opportunity set) erode the manager's alpha to zero. In their model, investors rationally chase performance (because recent outperformance signals skill), but the inflows eliminate the very outperformance that attracted them. The result is a market where skilled managers earn excess fees (the benefit accrues to the manager, not the investor) while investors earn index-like returns minus the higher active management fee. Hendricks, Patel, and Zeckhauser (1993) found short-term 'hot hands' in mutual fund returns (persistence over 1-4 quarters) but Bollen and Busse (2005) showed this persistence disappeared once daily data was used, suggesting it was driven by stale pricing and microstructure noise rather than genuine skill. The bottom line for investors: any selection methodology based on past returns -- star ratings, track records, league tables, or personal experience -- is likely to select funds that have already exhausted their outperformance and are poised to mean-revert.