Objective research to aid investing decisions

Value Investing Strategy (Strategy Overview)

Allocations for March 2024 (Final)
Cash TLT LQD SPY

Momentum Investing Strategy (Strategy Overview)

Allocations for March 2024 (Final)
1st ETF 2nd ETF 3rd ETF

Individual Investor Equity Market Timing

| | Posted in: Individual Investing, Investing Expertise

Should investors believe that they can usefully time the stock market? If so, how big might “usefully” be? In their July 2014 paper entitled “Can Individual Investors Time Bubbles?”, Jussi Keppo, Tyler Shumway and Daniel Weagley investigate persistence in the ability of individual Finnish investors to time the stock market, with focus on timing of two bubbles/crashes. They measure investor timing performance by relating monthly flows into and out of the investor’s portfolio to next-month and next-quarter returns of the value-weighted HEX 25 Index (now the OMX Helsinki 25). They test for persistence by comparing an investor’s relative timing performance in the first half of the sample period (January 1995 through March 2002) to that in the second half (April 2002 through June 2009). They treat January 2000 and October 2007 as beginnings of market crashes and focus on whether an investor performed well during the 12 months before and after each peak. Using data on all trades by 1,386,540 individual Finnish investors during January 1995 through June 2009, they find that:

  •  Across all investors in aggregate, correlations between monthly portfolio flows and next-month stock market returns are -0.01 in first half and -0.01 in the second half of the sample period, indicating that investors in aggregate do not successfully time the market.
  • For investors active enough to compare first and second half performances, average correlations between monthly portfolio flows and next-month stock market returns are 0.03 in the first half and 0.00 in the second half.
  • Investors who successfully time the market during the first half of the sample period are more likely to time the market successfully in the second half, and investors who successfully time the bubble/crash around 2000 are more likely to time the bubble/crash around 2008 successfully. For example:
    • 24.4% (17.3%) of investors who are in the top fifth of monthly market timers during the first half are in the top (bottom) fifth during the second half.
    • 17.2% (22.7%) of investors who are in the bottom fifth of monthly market timers during the first half are in the top (bottom) fifth during the second half.
  • The fifth of investors with the best market timing performances during the first half of the sample period:
    • Outperforms the fifth with the worst by roughly 9%-10% per year on a gross basis during the second half.
    • Outperforms buying and holding the market by roughly 3% annually on a gross basis during the second half.
  • The fifth of investors who perform best during the 2000 bubble/crash outperforms the fifth who perform worst by roughly 6%-8% annually during the second half of the sample period.
  • Older investors tend to be better market timers than younger investors. Ages 45-64 are most likely to be in the top fifth of market timers, and ages 65+ are least likely to be in the bottom fifth. Investors who trade more and make larger trades tend to have lower timing ability.

In summary, evidence supports some belief that equity market timing ability exists and derives at least partly from investing experience.

Cautions regarding findings include:

  • Per the above, while evidence of persistence is statistically significant, findings involve substantial shifts in performance rankings for many investors from the first to the second half of the sample period, meaning that a very large sample is needed for confident inference.
  • Economic significance findings apparently do not account for trading frictions. Accounting for trading frictions would reduce second-half outperformance of the market by the first-half best timers.
Login
Daily Email Updates
Filter Research
  • Research Categories (select one or more)