Understanding a Mutual Fund’s Average Annual Return; The calculation behind these performance figures may not be what you think; Rolling Returns Tell More About Fund Performance

Understanding a Mutual Fund’s Average Annual Return

The calculation behind these performance figures may not be what you think

SIMON CONSTABLE

Feb. 3, 2014 4:21 p.m. ET

What does it mean when a mutual fund reports its average annual return over a period of, say, three or five years? It isn’t exactly what you might think. What’s more, computing the figure yourself will require more than the general arithmetic you use in other areas of your life.

Specifically, average annual return can’t be determined by calculating the simple average of three or five one-year returns—the way you would calculate the average height of two people who are 5 and 6 feet tall, respectively. That’s because investment returns are volatile and the results compound year after year, says Paul Justice, director of data methodology at Morningstar Inc. MORN -4.17%

Consider a person who invests $100 and has a 10% loss one year and a 10% gain the next. That might seem to be an average return of zero. But the investment would actually be down one dollar in value—dropping 10% to $90 the first year, then growing 10% to $99 the next.

So fund companies and data providers typically report multiyear returns as “compounded average annual returns,” or geometric returns, which reflect how a series of returns affect an initial investment. The annualized return in our example is about negative 0.5%.

The details of the calculation aren’t important because fund companies will do the sums for you.

What is important to remember is that even seemingly small differences in annual returns compound over many years to create big dollar differences in how an investment grows. If you started with $10,000 and invested it for a couple of decades, a difference in average annual return of, say, half a percentage point could easily mean a difference of a few thousand dollars at the end of the period.

Quoted fund returns generally include the reinvestment of dividends and capital-gains distributions. They are net of operating expenses, but typically not of any sales charges.

 

Rolling Returns Tell More About Fund Performance

Look at shorter, overlapping periods to augment the usual trailing returns

CHRIS GAY

Updated Feb. 3, 2014 4:22 p.m. ET

Experienced fund investors know past performance is a poor indicator of future performance. But many don’t realize it can also be a poor indicator of the past.

The problem centers around one of the most widely used performance indicators, known as the “trailing return.” Pick up a fund prospectus or business publication and you’ll see references to trailing returns, which provide basic snapshots of a fund’s recent history—how a fund performed, usually over one-, three- and five-year periods ending on the same date.

According to some investing experts, however, trailing returns don’t say nearly enough about a fund’s real long-term performance and, worse, sometimes send misleading signals, particularly to unsophisticated investors.

To really see what’s beneath the hood of a mutual fund, some experts recommend looking at “rolling returns,” which break a performance period into many smaller, overlapping periods—a bit like slowing a movie down to study it frame by frame. Some analysts, for example, will break a five-year period into a series of rolling 12-month periods: January of year 1 through December of year 1, February of year 1 through January of year 2, and so forth.

Seeing a five-year period as a series of 49 rolling 12-month returns can reveal the ups and downs along the way. Similarly, rolling returns, because they draw on so many more data points than trailing returns, can reveal important differences in funds even when the funds have similar cumulative returns.

More Detail

Rolling returns “give you a much better sense than any point-in-time data will of how a fund has performed over its history and how a manager has performed over his or her history,” says Daniel Wiener, chief executive of Adviser Investments, Newton, Mass., which manages $3 billion in mutual-fund assets.

Morningstar.com is one place where investors can find free rolling-return data. Select the “Chart” tab on a fund’s main page, then the “Growth” pull-down menu at the top left of the chart to switch to “Rolling Returns.”

Consider the Vanguard 500 Index Fund, which tracks the S&P 500. According to Morningstar Inc., the fund posted a compounded annual return of 6.7% in the 10 years through Jan. 31. But looking only at that single-point figure, you wouldn’t know that the decade included a plunge of 37% in 2008, or a 32.2% gain last year.

Even those two numbers mask some of the fund’s volatility: Rolling returns show that the sharpest single 12-month decline was 43.3% (through March 2009) and the steepest 12-month rise was 53.6% (through March 2010). The average of the fund’s 109 12-month returns over the decade through December was 8.1%.

Avoiding Errors

How does it help to see the longer term in more granular detail? One big reason is investors who think about the ups and downs along the way—or who know a fund’s average returns for many periods, compared with a return based on a single starting and ending point—are likely to keep cooler heads in times of market volatility. Basing a decision to sell on a recent plunge, for example, is what sent many people fleeing equities in 2008-09 when they should have been getting in.

Similarly, many investors today are piling back into equities based on recent performance at a time when prices are near record highs.

“I think these big numbers [in recent trailing returns for equities] are part of the reason that you’re seeing this switch in flows out of bond funds and into stock funds,” says Mr. Wiener. “People are using [single] point-in-time numbers, looking at the S&P, and saying, ‘We’re getting 19% over five years! That’s fantastic!’ ”

Mr. Wiener and others note that trailing five-year returns for equities are likely to look even more tempting in the next few months, as the final months of the 2007-09 bear market drop out of the calculation.

“If you took a look at any of those funds [on Morningstar] in March 2009 and asked how they’d done over the past five years, it would look pretty sad,” says Doug Short, vice president of research for financial-data provider Advisor Perspectives in Lexington, Mass. “It looks a whole lot better now, especially if you’re looking at a five-year return.”

The sharper focus of rolling returns on multiple smaller periods also can be helpful in assessing performance of a fund or portfolio in which there is steady turnover in the pool of assets. In such cases, looking at returns for a rolling series of periods can be much more revealing of how the underlying assets contributed to overall performance than simply knowing what the fund’s returns were after one, three and five years.

Michael Herbst, director of active-funds research for Morningstar, says if he wants to evaluate a manager of a fund in which the typical asset is sold after five years, “I’m going to look at [the fund’s] rolling five-year returns, because that’s a time horizon that’s more consistent with the time horizon of the manager.”

Some Limitations

To be sure, rolling returns have their limitations. For one, to make them statistically significant, you need lots of historical data to draw from. That might not be available for newer funds. Similarly, older data isn’t relevant if a fund had a management change only a few years ago, Mr. Herbst says.

Another drawback: Averaging a series of rolling returns tends to overstate the performance of managers who do well in the middle of a rolling sequence—say, years 2, 3 and 4 of a five-year sequence—even though they did no better overall than managers who performed best in years 1 or 5.

“That’s a weakness in that it’s not truly the average that you might think it is,” says Ted Ponko, senior product analyst for Klein Decisions, a Research Triangle Park, N.C., maker of software for investment professionals. “Those periods in the first 12 months and those periods in the last 12 months aren’t being weighted the 12 times that all the other ones are.”