Financial advisors should know better than to chase the latest hot investment, and a research report offers more evidence that it’s a bad idea.

Equity-focused exchange-traded funds that saw big inflows as a percentage of assets under management over a short time period generally underperformed funds with average inflows, according to the “Shun ETFs With Largest Inflows” report by Jun Zhu, a co-portfolio manager for the Leuthold Core Investment Fund and a senior analyst at the Leuthold Group.

With more than 2,000 ETFs available, and many having multiyear histories, Zhu applied the back-testing methodology Leuthold commonly uses for stocks. Although ETFs debuted in the U.S. in 1993, they didn’t start attaining critical mass until about 2006 when the number of U.S.-listed ETFs first exceeded 200.

Zhu looked at fund flow data from April 2006 to April 2018 for all ETFs to assess the performance relationship. The lion’s share of ETFs are equity (69%), followed by fixed income (16%), commodities (8%), alternatives and asset allocation (both 3%), and currency (2%).

She split the ETFs into five groups using one-month, two-month and three-month fund flows as a percentage of month-end assets under management, and calculated the equal-weighted returns of each basket. The first quintile saw the largest inflows, the second quintile saw the second-largest inflows, etc. The last quintile saw the largest outflows.

The forward one-month return for ETFs with the highest one-month fund inflows underperformed ETFs in the other four quintiles. The highest inflow ETFs saw a negative 0.01% while the rest of the funds saw a range of positive returns from 0.37% to 0.48%. Zhu says this pattern persisted in back tests based on two-month and three-month fund flow data.

She also looked at how the fund-flow effect and performance appeared over six-month, nine-month and 12-month periods, and this also showed funds with the biggest inflows underperformed over all return horizons.

Over the course of the 12-year study period, people who invested equally weighted in equity ETFs with the highest inflows saw an 8% loss. Meanwhile, an equal-weighted portfolio using the other four ETF baskets returned 72% during the same period.

While Zhu can’t say exactly why this happens, she has a hunch.

“We know equity investors, especially with retail investors, have the tendency to chase returns,” she says.

That herd behavior tends to swell the valuation of the underlying assets in the ETFs, but eventually that valuation will revert back to normal and the asset will eventually underperform.

That pattern in the equity ETFs, however, is missing from fixed-income ETFs, Zhu notes. The inflows and performance across the quintiles in the fixed-income funds has been flat since 2009.

“Fixed-income investors are more risk averse. They want to protect their principal or protect their investment. So instead of return-chasing, when things are not going well they may overreact [and sell],” Zhu says.

There aren’t enough ETFs investing in commodities and currencies, Zhu says, to make a meaningful analysis. She says she’ll need to look closer to ascertain why equity ETFs with high inflows relative to AUM underperformed, but for now she says it’s another reminder to be thoughtful when investing.

“It’s just something that you want to keep in mind when you are trying to build your portfolio, to probably avoid certain ETFs which are really hot,” she says.