Should you pay attention to Morningstar’s star ratings? No, say a growing number of advisors. The stars are based on risk-adjusted past performance, and the past doesn’t necessarily predict the future, the advisors say. Instead of trying to pick tomorrow’s winning active managers, it pays to rely on index funds.
Morningstar concedes that its system is not foolproof, but the company has long argued that the stars can provide some guidance. Now Morningstar is conducting a new research project, aiming to determine more precisely what the stars can do. Besides looking at the star system, the researchers are looking at Sharpe ratios and other measures of past performance. The goal is to determine whether using combinations of indicators can result in more reliable predictions of future performance.
John Rekenthaler, Morningstar’s vice president of research, discussed the preliminary findings at the Morningstar Ibbotson conference that was held recently in Orlando. So far the new research demonstrates that the star ratings closely track Sharpe ratios, which measure risk-adjusted returns. Both measures are useful to consider Rekenthaler said. “Risk-adjusted performance appears to be mildly predictive of future performance,” he said.
For the study, Rekenthaler looked at how stock and bond funds performed over 5-year periods. On average, funds that had the top rating of five stars went on to return 7.3 percent annually during the next five years, while lowly 1-star funds returned 6.5 percent. The 5-star funds delivered higher Sharpe ratios. The system worked better for certain categories. While 5-star balanced funds returned 11.1 percent during the five years ending in 2007, 1-star funds only returned 8.1 percent. But the system was far from perfect. During some periods, top-rated funds trailed those with fewer stars. For example, 5-star domestic stock funds returned 2.9 percent annually during the five years starting in January 2006, while 4-star funds returned 3.0 percent.
Though the five-star funds had a mixed record, it is important to keep in mind that the data on average performance only includes funds with five-year records. Funds that went out of business before the period ended were dropped from the data base. When fund liquidations are considered, the 5-star funds appeared clearly superior. Of the 1-star domestic stock funds that existed in January 2006, 55 percent shut their doors during the next five years. Only 9 percent of 5-star funds merged or liquidated, and 18 percent of four-star ones closed.
The data on fund terminations suggests that the stars are worth considering because no one wants to invest in a fund that merges or goes out of business. When funds vanish, clients can be annoyed at the advisors who recommended the losers.
Besides looking at the star system, Morningstar also examined the impact of fees. It has long been known that funds with low expense ratios tend to excel in the future, and the Morningstar study confirmed this. During the five years that began in 2006, stock and bond funds with expense ratios in the cheapest 20 percent of their categories returned 4.3 percent, compared to 3.1 percent for funds in the most expensive quintile. Expense ratios appeared to be a slightly better predictor of future results than the star system. However, the stars did a better job of helping investors to avoid funds that went out of business. Of 5-star funds, only 8 percent went out of business during a five-year period, while 14 percent of cheap funds merged or liquidated.
Rekenthaler found that the results could be improved by considering both stars and expenses. During the five-year period that began in 2006, five-star taxable bond funds returned 5.5 percent, while funds in the cheapest quintile returned 5.6 percent. Of funds that were in the cheapest quintile and had five stars, the returns were 6.0 percent.
Morningstar is continuing to look for indicators that would be superior to the star system. One promising technique is modifying the star system to consider a manager’s tenure. This is different from the traditional star system, which makes no effort to track manager changes. Under the current system, a fund can have five stars, even though the current manager has been at the helm for two months. In the modified system, the fund would only be rated for the period when the current manager was in charge. Preliminary research suggests that evaluating manager track records is superior to the traditional star ratings