Investors’ choice of benchmark is “not statistically significant,” according to a long-term study of index performance.
“We conclude that the strongest driving force of considered benchmark portfolios is the market factor.”Having tracked the performance of S&P 500 stocks over a 23-year period from 1989 to 2011, Yuliya Plyakha of the University of Luxembourg claimed that “researchers should not worry too much about the choice of the benchmark portfolio for testing asset pricing models in case of sufficiently long sample period.”
“Value-weighted and equal-weighted portfolios of 500 stocks can be used interchangeably,” Plyakha stated in her paper, “Benchmarking Benchmarks: Much Ado About Nothing.”
Plyakha—who now works in MSCI’s portfolio analytics team—studied the performance of S&P 500 constituents according to a number of “traditional and new” benchmark rules, including equally weighted, minimum variance, Sharpe-ratio constrained, and dividend-based models.
“We conclude that the strongest driving force of considered benchmark portfolios is the market factor,” Plyakha wrote. More than two-thirds of the dynamics observed in the 23-year period covered were down to the “market systematic factor”—or market beta—the paper said.
“This result implies for big mutual, pension, and hedge funds with long investment time horizons that the choice of their benchmark is not important, and that sticking to traditional value-weighted benchmark is good enough,” Plyakha said. “If the fund is small, around 100 assets, equal-weighted benchmark would be a better choice.”
Plyakha added that her findings demonstrated that “irrespective of the benchmark portfolio, fund managers mainly track the market and do it in a statistically sufficient way.”