Morningstar's Kathryn Young discusses the pitfalls lurking in trailing returns data and the best ways to use it in fund assessment.
Trailing returns are the most basic and widely-used measure of a fund's historical performance, and with good reason.
They measure total returns over standardised time periods, usually with a recent ending date, making them easy to understand and readily comparable with a relevant index or peer group.
Three-year trailing returns have become a particularly important metric.
Industry participants tend to use that time period as a hurdle for admission onto platforms, and many investors have become comfortable relying on that track record for their decisions.
The problem, however, is that trailing returns are inherently time-period-sensitive, and three years isn't nearly long enough to provide a full picture of a fund's performance.
Reliance on trailing returns – especially over periods of three years and less – can lead to undesirable outcomes.
Recent performance updates offer a stark example of this issue.
Over the second half of 2011 and the first part of 2012, market losses stemming from the global financial crisis cycled out of three-year trailing return calculations.
The contrast between the deep lows experienced during the crisis and the extreme highs during the recovery that began in March 2009 has led to dramatic fluctuations in trailing returns from month to month, as the data for the funds in Table 1 demonstrates.
Note the significant increase between the returns figures for the periods that started in October 2008 and November 2008.
Australian share funds generally suffered extreme losses in September and October 2008 directly relating to the Lehman Brothers bankruptcy, so their three-year returns perked up substantially after those months rolled off.
Similarly, the returns beginning in February 2009 and March 2009 reflected significant increases as the losses in January and February 2009 rolled out of the calculation.
The funds’ three-year category rankings (in brackets after each return figure) and comparison to the relevant index demonstrate that these dramatic return fluctuations have had a powerful impact on the funds’ relative standing.
For example, Perennial Value Australian Shares 6821 generally fared well through the crisis relative to its Australian Large Value peers, but lagged them as the rebound started. As a result, its category rank declined for the period starting March 2009.
In contrast, Lazard Australian Equity Wholesale 10700 lost substantially more than the average large-value fund in January and February 2009, so its category rank improved considerably as that performance rolled out of its three-year return rank calculation.
These dramatic shifts in funds’ relative standings can have a major impact on investment decisions.
Investors who use these performance measures in isolation might, for example, select Perennial Value if using October 2011 data, but Lazard if looking to invest in April 2012.
The two funds are quite different, and are likely to produce divergent investment experiences. Perennial Value, which we rate Silver, employs a relative value approach, is benchmark-aware, and tends to perform well in rising markets.
Neutral-rated Lazard applies a deep value strategy and has historically looked very different than the S&P/ASX200 Index thanks to a persistent underweight in materials stocks.
Despite its underperformance in early 2009, Lazard typically fares better in a falling market than Perennial Value, but lags in rallies.
This demonstrates that reliance on trailing returns – especially over periods of three years and less – could lead to undesirable outcomes.
This phenomenon isn’t exclusive to Australian or equities assets.
The funds in Table 2 confirm that while the effect may be different depending on each asset class’ performance during the extreme turbulence of the global financial crisis, the time period sensitivity of trailing return and relative ranking calculations persists across different asset classes.
The issues demonstrated here don't make trailing returns useless.
They remain valuable for their convenient, straightforward characteristics. The drawbacks, however, are important to keep in mind.
Trailing returns should be evaluated in the context of market conditions and fund strategy.
For example, funds with more defensive, conservative strategies tended to fare better during the global financial crisis, so their returns look relatively worse as that period rolls off. Similarly, funds with more aggressive strategies can look better over periods of more sanguine market conditions.
Our research reports offer a description of a fund’s strategy and performance tendencies. In addition, the flaws inherent in trailing returns underscore the importance of evaluating funds over the longest applicable time period available.
Five and 10-year trailing returns have been less affected by the fluctuations discussed above, and 10-year returns should comprise a full market cycle, providing a picture of the fund's performance throughout a variety of market conditions.
That principle applies when reviewing the existing pieces of a portfolio as well.
It’s best to resist the urge to make changes on the basis of short-term performance data, and to instead take a longer-term perspective and a holistic approach to analysis that also includes qualitative elements.
Kathryn Young is a fund research analyst at Morningstar.