Objective research to aid investing decisions

Value Investing Strategy (Strategy Overview)

Allocations for January 2021 (Final)

Momentum Investing Strategy (Strategy Overview)

Allocations for January 2021 (Final)
1st ETF 2nd ETF 3rd ETF

Two Self-destructive Individual Investor Behaviors

| | Posted in: Animal Spirits, Individual Investing

What individual investment behaviors are worst? In their January 2014 paper entitled “Which Investment Behaviors Really Matter for Individual Investors?”, Joachim Weber, Steffen Meyer, Benjamin Loos and Andreas Hackethal investigate relationships between the following ten tendencies of individual investors and portfolio performance:

  1. Portfolio turnover: unprogrammed trading volume scaled by portfolio value.
  2. Trade clustering: clustering of investor trades in time.
  3. Disposition effect: selling of winners and holding of losers.
  4. Leading turnover: trading before other investors (same security/same direction).
  5. Forecasting skill: systematically realizing excess returns on purchased securities.
  6. Trend following: buying funds with recent increases in value.
  7. Home bias: preference for German stocks or Germany-focused funds.
  8. Local bias: preference for stocks/funds with nearby headquarters.
  9. Lottery mentality: preference for stocks with low price and high idiosyncratic volatility/skewness.
  10. Under-diversification: holding only a few securities and/or highly correlated securities.

Using trading records, monthly position statements and demographics for 5,000 predominantly German individual investors who use a discount broker spanning January 1999 through November 2011, they find that:

  • Most portfolios concentrate in stocks and mutual funds, with an average of 11 positions. For the entire sample, the average (median) number of portfolio transactions per year is 36 (16).
  • For the entire sample, average monthly gross (net) portfolio return in excess of the risk-free rate is -0.47% (-0.59%).
    • Average monthly net excess return for the stocks (funds) only in portfolios is -0.68% (-0.24%).
    • Average net four-factor (market, size, book-to-market, momentum) alphas are slightly negative for overall portfolios and stock holdings, and about zero for fund holdings. Dependence of holdings on the four factors has about twice the power to explain returns as investor behaviors.
  • Single-variable analyses indicate that portfolio turnover, trade clustering, lottery mentality and under-diversification hurt returns, while forecasting skill and trend following help returns.
  • However, multi-variable analyses that account for variable interactions indicate that only two behaviors, under-diversification and lottery mentality, significantly affect returns. Both effects are negative.
    • Eliminating under-diversification and lottery mentality behaviors would boost average annual net portfolio return by 3.6% and 3.3%, respectively, and would further improve risk-adjusted performance since both behaviors relate positively to idiosyncratic portfolio risk.
    • Findings are robust to investor demographics and different performance measures/benchmarks. They hold for full portfolios and stock holdings only, but not for fund holdings only.
  • Portfolios of the fifth of investors making the fewest behavioral mistakes across all ten behaviors outperforms portfolios of the fifth making the most mistakes by an average gross 8.1% per year.

In summary, evidence from a large sample of German investors indicates that under-diversification and lottery mentality are the two most self-destructive behaviors of individual investors.

The authors suggest an easy way to avoid one of the two significantly self-destructive behaviors: “Under-diversification could simply be reduced by purchasing products that offer built-in diversification, such as exchange-traded funds or mutual funds.”

Cautions regarding findings include:

  • As noted in the paper, many behavior-return relationships are non-linear, with significance concentrated in behavioral extremes.
  • Given the source of data, gross minus net in this case is due only to broker fees. The implied level of trading friction is therefore not comparable to that based on models that use closing prices with bid-ask spreads plus broker fees to account for trading friction.
Daily Email Updates
Filter Research
  • Research Categories (select one or more)