Morningstar's New Metric: "Investor Return"

eginning in October 2006, Morningstar introduced a new measure of performance called Investor Return. It is a dollar-weighted measure of return, whereas total return is a time-weighted measure. Both are annualized returns when applied to multi-year periods.

"The new measure, called Morningstar® Investor Return(TM), estimates the return earned collectively by all the investors in a fund. Investor return, also known as dollar-weighted return, accounts for all cash inflows and outflows from purchases and sales and the growth in fund assets. It complements the more traditional metric of total return, which measures what investors could have earned had they bought and held the fund, reinvesting all dividends, over a period of time." (Morningstar)

"Investors know they should hold diversified portfolios, but many chase past performance and end up buying funds too late or selling too soon," said Don Phillips, managing director of Morningstar. "For example, if most investors bought a fund at a high point and sold in a trough, that fund's investor return would be lower than its total return. Similarly, a fund's investor return might be higher than its total return if investors, in aggregate, timed their purchases and sales astutely. In essence, investor return shows how well-or how poorly-investors in a fund have fared."

Understandably, this new return metric is greatly influenced by timing of asset flows in and out of a fund, which may not be an accurate reflection of the quality of the fund, but rather the behavior of impulsive investors and/or advisors. In light of that, Phillips goes on to say, "Fund companies don't have complete control over how investors use their funds, but that doesn't mean they can't exercise any control. Fund companies can influence investor behavior through fund design, the timing of launches and closings, marketing efforts, and shareholder communications. With access to investor returns, investors and advisors can determine which fund companies are successfully deploying strategies that result in a better outcome for investors."

This study attempts to uncover mutual fund performance characteristics that are related to dollar-weighted performance in an attempt to decipher what Investor Return tells us and what it doesn't. The data for this study were extracted from the October 2006 release of Morningstar Principia. Only U.S. equity funds were considered. Redundant funds were removed (meaning that only share class of a multiple share class was retained). Funds included in this study had a 10-year performance history as of 9/30/06 and an equity style consistency rating of "High" (to minimize the impact of style drift). A total of 563 funds met the criteria of this study.

Inasmuch as a wide variety of funds (covering all nine style boxes) were included in this study it was necessary to analyze the relationship between Investor Return (dollar-weighted) and the normal total return (time-weighted) rather than solely focusing on Investor Return. This was accomplished by calculating the ratio between Investor Return and total return. The advantage of calculating this ratio is that it creates a number that is independent of absolute levels of return, which allows for an "apples to apples" comparison of funds despite differences in market cap, equity style, or level of absolute total return.

In terms of total return, value funds generally outperformed growth funds during the 10-year period of this study (Sept 1996 - Sept 2006). Hence, the Investor Return of value funds tended to be higher than the Investor Return of growth funds. However, by calculating the ratio between Investor Return and total return (what I have termed "Investor Return Yield" or IRY) the raw total differentials between value and growth, large cap and small cap, etc. are removed. The formula of the ratio is as follows:

Investor Return (%)
------------------------		=	Investor Return Yield  (or IRY)
Total Return (%)

Investor Return Yield (IRY) is a measure of the ratio between Investor Return and the total return. Said differently, IRY calculates what percent the dollar-weighted return is of the time-weighted return. The higher the IRY the better. For instance, an IRY of 1.10 indicates that the returns of the fund — after taking into account the cash flows in and out of the fund-resulted in the "average investor" experiencing a return that was 10% better than the reported time-weighted total return.

As shown in "A Tale of Three Funds", the ratios between Investor Return and total return can be widely different. Each of these three funds is categorized by Morningstar as being in the Large Cap Growth style box. The IRY is a portable statistic which allows for meaningful comparisons between funds within the same style box or not. Over the 10-year period ending in September, 2006 the RSI Retirement Core Equity fund had an IRY of 1.75 indicating that its dollar-weighted Investor Return of 8.82% was 75% higher than its time-weighted total return of 5.05%. American Century Growth's Investor Return was 7% higher than its reported total return, while Vanguard Growth Equity had an IRY of -1.07 is a result of its Investor Return (-6.05%) being 207% lower than its 10-year annualized return of 5.67%. (IRY of -1.07 = % difference of 207%)

A Tale of Three Funds

As previously noted, Investor Return is heavily influenced by the fluctuations in a fund's net assets and the fund's subsequent performance. The data driving the IRY of these three particular funds are shown in "The Flow". RSI Retirement Core Equity had a fairly consistent asset base during the buoyant years of 1996-1999 and a slightly smaller asset base during the tough years of 2000-2002. Conversely, the assets in Vanguard Growth Equity were relatively small during 1996-1999 and then surged in 2000 as performance started into a three year decline. Thus, a relatively small percentage of the Vanguard Growth Equity funds' assets enjoyed the good years of 1996-1999 while vastly more assets experienced the negative returns of 2000-2002. The correlation between asset growth and subsequent performance was similar for American Century Growth, though not as dramatic. These three funds are used to illustrate the relationship between asset fluctuation and performance-the driving forces behind the calculation of Investor Return. Mention of these funds does not imply an endorsement or criticism.

The Flow

Data shown in "The Flow" are shown graphically in "Three Paths". The superimposed red line is the linear relationship between annual asset totals and annual total returns. Funds with higher Investor Return will have a positively sloped line. In fact, the relationship between the two can be described by a correlation coefficient.

The 10-year correlation of RSI Retirement Core Equity was 0.55 (where 1.00 equals a perfect positive correlation). This indicates a positive relationship between the level of net assets and annual performance-the key to a higher Investor Return and IRY. Most of the higher returns of the RSI fund occurred when asset levels were higher (relatively speaking). The R-squared between assets and performance was .30 (where 1.0 is the maximum) indicating that the predictive relationship between level of assets and performance was relatively low (only 30% of the change in one is explained by the other).

The 10-year correlation between assets and returns for American Century Growth was 0.25 (R-squared = 0.06). The Vanguard Growth Equity correlation was -0.57 (R-squared = 0.32). A negative correlation between assets and performance indicates that some of the lower returns occurred when asset levels were higher-the "key" to both lower Investor Return and lower IRY.

Three Paths

Having illustrated how Investor Return is calculated, we now turn our attention to IRY among this group of 563 funds.

The average 10-year IRY for the 563 funds in this study was 0.82-indicating that the average investor realized 82% of the stated total return. By comparison, the median IRY among the 563 funds was 0.92, which is more encouraging. The average IRY for the funds in each style box is provided in "Inside the Boxes".

Inside the Boxes

Value and blend funds fared very differently than growth funds over the 10 year period ending September 30, 2006. The cause is shown in "Rise and Fall". In 1996 and 1997 U.S. equity funds were enjoying high returns nearly across the board (value, blend, growth). As a result, assets were growing because of (1) high positive returns and (2) large net inflows. In both 1998 and 1999 cash continued to flow into growth funds (43% average increase in net assets in 1998 and 59% average increase in 1999) in response to their buoyant returns. Conversely, the performance of value funds and blend funds in 1998 and 1999 tapered off which slowed down the asset inflow. The stage was set for IRY to get nailed. Asset-heavy growth funds got clobbered in 2000, 2001, and 2002 whereas value and blend funds took a smaller beating as a result of having relatively fewer "late" assets. The term "late assets" refers to assets that flowed into growth funds near the end of the bull market (late 1999). The dynamic at work here is somewhat akin to Last-In-First-Out. However, in this context, it is more like Last-In-Most-Hurt. Viewed in this way, Investor Return (and IRY) are metrics for measuring the cost of chasing performance.

Rise and Fall

The 10-year IRY for value and blend funds (.91 and .92) was statistically higher than the IRY of growth funds (.65) at the 99% confidence level. Let's break this down further. The average annualized growth rate in assets for growth funds during the period from 1996-1999 was 51%, statistically higher than the 38% growth rate in assets among value funds (90% confidence level). During the same period, growth funds had an average annualized return of 24.1%, which was statistically higher than the return of value funds (14.5%) and blend funds (18.5%) at the 99% confidence level. Thus, growth fund buckets were filling up with assets just before they went into a three-year performance decline from 2000-2002. In fact, growth funds had an average annualized return of -13.2% during those three years, compared to -7.1% for blend funds and -1.2% for value funds (99% confidence level in both cases). Growth fund story: big asset build up from 1996-1999 + big performance dive in 2000-2002 = low IRY.

From 2003 through September 2006, the annual increase in net assets among value funds and blend funds was significantly more rapid (24% average annual rate) than among growth funds (9% average annual rate) in spite of the fact that the return of growth funds and blend funds had comparable performance. Value funds, on average, significantly outperformed blend and growth funds from 2003-2006. It would appear that investors are entering back into growth funds cautiously as a result of recent history.

The final table ("Differences in Growth") focused solely on the 126 large cap growth funds in the sample. The 126 funds were ranked by 10-year IRY in descending order. The 32 funds with the highest IRY formed the top quartile and the 32 funds with the lowest IRY comprised the bottom quartile. The averages for the top and bottom quartile were then compared and the difference in the means was tested for significance.

As would be expected, the 10-year IRY was significantly different (99% confidence level). The top quartile mean was 1.35 compared to a mean of -.40 among the bottom quartile funds. This indicates that high IRY large cap growth funds had an Investor Return that was an average of 35% higher than total return. Conversely, the bottom quartile funds had an average IRY of -.40 indicating an Investor Return lower than total return by 140% on average. (FYI, the median top quartile IRY was 1.29 and the bottom quartile median IRY was 0.17. Using median figures, the top quartile funds had an Investor Return that was 29% higher than total return, while bottom quartile funds had an Investor Return that was 83% lower than total return).

Interestingly, the 10-year annualized total (time-weighted) return was not significantly different between the top and bottom quartiles whereas the average 10-year Investor Return was different among top and bottom quartiles at the 99% confidence level. Therefore, we can clearly distinguish that it was not differences in raw total return that contributed to differences in IRY and Investor Return among these large cap growth funds. What was linked to low IRY?

We see that the correlation between annual asset levels and annual returns was 0.31 among the top quartile funds. This is a positive correlation indicating that as one variable goes up the other variable also tends to go up. In this context, when assets and returns go up together Investor Return tends to increase. Among the bottom quartile funds, there was a slight negative correlation (-.16) suggesting that assets and returns do not always move in the same direction. The difference between these two correlations was significant at the 99% confidence level. Bottom quartile funds also had significantly higher volatility of return over this 10-year period, as noted by a standard deviation of 19.9% compared to 15.7% among the top quartile funds. Moreover, bottom quartile funds had significantly higher average annual turnover ratio over the 10 years (93.6% vs. 67.5%). Average annual expense ratios tended to be slightly higher among the bottom quartile funds, but the difference was not statistically significant. Finally, the average annual percentage change in assets among the top and bottom quartiles was highly statistically different only during the 1996-1999 period. This last observation coincides with the general finding in "Rise and Fall".

Over this particular 10-year period we observe that large cap growth funds which had higher return volatility (as measured by standard deviation), higher turnover, and whose assets grew fast and late (near the end of the market bubble) were those funds with lower Investor Return and lower IRY. Raw total return and expense ratios were not correlated to inferior Investor Return.

Differences in Growth

In summary, Morningstar's Investor Return is a metric moving in the right direction. Accounting for asset flows in the measurement of performance is a realistic issue. However, a word of caution about Investor Return. It is very sensitive to the accuracy of net asset data from year to year. Also, Investor Return is most useful when comparing funds that have a high degree of similarity. To compare Investor Return among non-homogenous funds, it will be necessary to calculate the investor return yield or IRY.

____________________________________________________________________________________
Craig L. Israelsen, Ph.D. is an associate professor at Brigham Young University. He teaches family finance in the Department of Home and Family Living. His research interests include mutual fund analysis. He writes monthly for Financial Planning magazine. Learn more about Craig Israelsen at http://familyliving.familylife.byu.edu/faculty/israelsen.htm