Taking Issue with the Active vs. Passive Debate. Craig L. Israelsen, Ph.D. Brigham Young University. June Contact Information:

Size: px

Start display at page:

Download "Taking Issue with the Active vs. Passive Debate. Craig L. Israelsen, Ph.D. Brigham Young University. June Contact Information:"

Jeffry Crawford
6 years ago
Views:

1 Taking Issue with the Active vs. Passive Debate by Craig L. Israelsen, Ph.D. Brigham Young University June 2005 Contact Information: Craig L. Israelsen 2055 JFSB Brigham Young University Provo, Utah (phone) (fax) Craig L. Israelsen, Ph.D., is an Associate Professor in the Department of Home and Family Living at Brigham Young University where he teaches Family Finance. He holds a Ph.D. in Family Resource Management from Brigham Young University. He received a B.S. in Agribusiness (1983) and a M.S. in Agricultural Economics (1984) from Utah State University. Primary among his research interests is the analysis of mutual funds. He writes monthly for Financial Planning magazine. 1

2 Abstract The active vs. passive investing debate will continue on. This study suggests two refinements when making such comparisons: (1) calculate and use aggregate index returns rather than individual benchmark index returns, (2) count the amount of outperforming assets rather than the number of outperforming funds. 2

3 Taking Issue with the Active vs. Passive Debate Few topics in the field of finance have generated the interest and spirited debate as has the issue of active versus passive investing (Grinold and Kahn, 1999; MacKinlay, 1998). Empirical evidence in support of the superiority of passively managed portfolios is persuasive (Davis, 2001; Arnott, Berkin and Ye, 2000; Sorensen, Miller and Samak, 1998; Carhart, 1997; Gruber, 1996; Malkiel, 1995; Brinson, Hood and Beebower, 1995). Conversely, equally sound empirical and/or logical evidence in defense of the value of active portfolio management has been presented (Pastor and Stambaugh, 2002; Wermers, 2000; Elton, Gruber and Blake, 1996; Etzioni, 1992). Moreover, behavioral arguments in favor of actively managed funds have been offered (Timbers, 1997). This study addresses, and attempts to resolve, several empirical issues that commonly impinge on the analysis of active vs. passive investing, specifically in relation to actively or passively managed mutual fund portfolios. First issue: Passive investing in the U.S. equity market is generally equated with investing in equity-based index funds. The performance of such an approach is too often represented by use of a single index the Standard & Poor s 500 Index. Representing the broad U.S. equity market by use of one large-cap equity index dramatically simplifies (and distorts) the complexity of the U.S. market, or virtually any equity market for that matter. Rather than assume that the U.S. equity market is represented by a single index (say, the S&P 500) this study calculates the average return of major U.S. equity indexes within each of the nine Morningstar style boxes (large cap value, large cap blend, large cap growth, mid cap value, mid cap blend, mid cap growth, small cap value, small cap blend, and small cap growth). 3

4 The averaged returns of several major equity indexes are then used as the performance level against which actively managed funds within the same style box are compared. Second issue: Most, if not all, active vs. passive investing studies tally the number of funds that beat their respective benchmark over some specific time frame. This approach has the unfortunate effect of equally weighting funds. Clearly, however, a fund with $30 billion in assets that outperforms its benchmark index is more important than an out-performing fund with only $30 million in net assets. It must be agreed that when studying this issue the ultimate measure of relevance is how the end-user the investor is impacted. As such, measuring the asset total that outperforms this or that index is the more correct measure of impact. Therefore, rather than count the number of funds which over or under-perform a particular equity benchmark, this study summed the total mutual fund assets (within each distinct style box) that outperformed the performance of their comparable benchmark indexes. Importantly, this approach also resolves an additional issue, namely what to do with mutual funds that are sold via multiple share classes. The removal of redundant share classes is important in many types of mutual fund analysis so as to avoid the problem of over-weighting the performance of funds with multiple share classes. However, as this study is seeking to know how many dollars outperformed a target benchmark, it doesn t matter if the assets were in A shares, B shares, C shares, etc. All that matters to the individual investor is how their fund (regardless of share class) performed relative to appropriate benchmark(s). Data The first task was developing aggregated, index-based performance benchmarks against which the performance of actively managed equity mutual funds can be appropriately measured. Performance data from six different index makers were utilized in this part of the study, 4

5 including Barra, Dow Jones, Russell, Wilshire, Morningstar, and Standard & Poor s. Performance data for U.S. equity indexes were obtained from Morningstar s Principia software, specifically the January 2003 release, January 2004 release, and the January 2005 release. The aggregated performance of five different large cap value U.S. equity indexes is reported in Table 1. The benefit in calculating the average return for the various indexes is manifested by the differential between the high and low return in each of the three years. In 2002, for instance, there was a 534 basis point differential between the best performing large cap value index (Russell 1000 Value) and the worst performing index (Barra Large Cap Value). A performance differential of that magnitude will have a sizable impact in a comparison of actively managed funds against passive indexes based upon which index is chosen as the benchmark. The averaged results of mid cap indexes (value, blend, and growth) for 2002, 2003, and 2004 are reported in Table 2. A difference of 534 basis points among large value indexes in 2002 is a small thing compared to the difference observed between the best and worst performing small growth index in As shown in Table 3, there was a 2,353 basis point differential between the Barra Small Cap Growth Index and the Dow Jones Small Growth Index. The argument for or against active management will look very different based upon which Small Growth Index is used as the benchmark. A summary of the average differential (in basis points) between best and worst performing equity index over the three year period from 2002 to 2004 within each Morningstar style box is provided in Table 4. Blend indexes have the lowest average discrepancy between high and low index, followed by value, and then growth. Large cap indexes have less discrepancy than mid cap indexes. Small cap indexes have the most variation, particularly so among small cap growth indexes. The average annual difference between best and worst 5

6 performing small cap growth index was in excess of 1,400 basis points over this particular three year period. Based upon the degree of discrepancy between major U.S. equity indexes there is a clear need for aggregating performance of individual indexes and creating an average benchmark return as part of any active versus passive investing analysis. On to the second issue, that of measuring outperforming assets, rather than counting the number of outperforming funds. Data for this portion of the analysis were obtained from three separate year-end Morningstar Principia databases (year-end 2002, year-end 2003, and year-end 2004). Use of each year-end database to gather annual mutual fund performance, as opposed to using only the year-end 2004 Principia database to gather three year s worth of performance data minimized the impact of survivorship bias. Moreover, using each year-end database allowed for retrieval of net assets at the end of each year. Actively managed U.S. equity mutual funds included in this analysis met the following criteria at the end of each of the three years: each fund had to be classified as being in the Domestic Stock Morningstar category, had to have a best-fit beta greater than or equal to 0.50, had to have at least 66% of their portfolio in U.S. stocks, and less than 25% of its portfolio in non-u.s. stocks, bonds, cash, or other. Index funds were, of course, omitted. Finally, there was a unique filter for each sub-group based upon the Morningstar Style Box (i.e. funds with an equity style box of Large Value were selected separately, then Large Blend, and so forth). Mutual fund returns were not adjusted for taxes or sales loads. Sales loads are a tricky issue inasmuch as some index funds assess a load, hence it is unfair to assess a sales load on the actively managed fund and not assess a load against the return of the indexes. As it is, a comparison of actively managed funds against the performance of indexes is already slanted in favor of indexes. Indexes do not have expense ratios, but actively managed funds do. As raw equity indexes are not purchasable products, 6

7 investors can only gain access to indexes via index funds which have expense ratios. However, the expense ratio of the actionable, purchasable index funds is ignored when using the returns of raw indexes as the performance benchmark against which actively managed funds are compared. Hence, using raw equity index performance (averaged performance in this study) as a benchmark against which actively managed funds are compared is inequitable from the start. Nevertheless, it is a common practice and, as such, was followed in this study. Results As shown in Table 5, there were 590 actively managed large cap value funds at the end of Of that total, 243 funds (or 41.2% of the group) had a one-year return greater than the average large cap value index return of -18.0% (as reported in Table 1). However, of the $299.9 billion assets classified as residing in the large cap value style box at the end of 2002, $187.3 billion were in funds with a one-year return in excess of the average large value benchmark return of -18.0%. Therefore, 62.4% of large value assets beat the average benchmark return, whereas 41.2% of large value funds beat the benchmark average return. In 2002, six out of nine style boxes reported a higher percentage of outperforming assets than outperforming funds. If all nine style boxes are combined to represent the entire U.S. equity market, nearly 54% of all actively managed mutual fund assets beat their respective average index return. Conversely, of the 4,075 funds at year-end 2002, only 1,795, or 44.1%, exceeded the average return of their respective indexes. One plausible implication of this result is that in 2002 (a very difficult year for U.S. equities) funds with larger net asset bases performed better in comparison to their respective indexes than did smaller funds. The year 2003 was an excellent year for U.S. equities across the board. As shown in Table 6, the nearly 43% of actively managed fund assets beat their respective index returns, whereas 38% of funds did so. In 2004, nearly 51% of U.S. equity fund assets had a one-year 7

8 return of their respective benchmark indexes while less than 41% of actively managed funds beat their benchmarks (see Table 7). Among actively managed large cap growth funds in 2004, nearly 82% of their $686+ billion in assets outperformed the average large cap growth index return of 5.6% whereas only 65.4% of the large cap funds had a return in excess of 5.6%. In eight out of the nine style boxes the percentage of actively managed assets which beat their respective benchmark indexes exceeded the percentage of actively managed funds that beat their indexes. Summarized in Table 8 are the three year averages of total amount of assets and number of funds whose performance beat their benchmark indexes. In general, the discrepancy between percentage of assets and percentage of funds which beat their indexes is relatively small among value funds; increases among blend funds, and then sizably increases among growth funds. Likewise, among small cap funds the discrepancy is small and increases as one moves up in market capitalization. For instance, over the three year period being studied, the percentage of small cap assets that beat their benchmarks was 45.4%, whereas 43.4% of small cap funds beat their benchmark indexes. Conversely, 50% of large cap assets outperformed the indexes while only 40.8% of large cap funds did likewise. Summary When comparing the performance of actively managed funds against that of passively managed index funds, it is important to measure outperforming assets rather than outperforming funds. This is particularly true among large cap mutual funds and growth-oriented mutual funds. Small cap funds and value-oriented funds tend to have less discrepancy between percentage of outperforming assets and percentage of outperforming number of funds. (see Figure 1). 8

9 Another important element in the active vs. passive debate is calculating average index performance using several comparable indexes rather than using a single index as the performance barometer for any single style box. 9

10 Table 1. Large Cap U.S. Equity Indexes Yr. Annualized LARGE VALUE U.S. EQUITY INDEXES Barra Large Cap Value Dow Jones Large Value Russell 1000 Value Russell Top 200 Value Wilshire Large Value Average Large Value Index Performance Basis Point Differential Between High and Low LARGE BLEND U.S. EQUITY INDEXES Morningstar Large Cap Russell Standard & Poor's Wilshire Large Cap Average Large Blend Index Performance Basis Point Differential Between High and Low LARGE GROWTH U.S. EQUITY INDEXES Barra Large Cap Growth Dow Jones Large Growth Russell 1000 Growth Russell Top 200 Growth Wilshire Large Growth Average Large Growth Index Performance Basis Point Differential Between High and Low

11 Table 2. Mid Cap U.S. Equity Indexes Yr. Annualized MID VALUE U.S. EQUITY INDEXES Barra Mid Cap Value Dow Jones Mid Cap Value Russell Mid Cap Value Wilshire Mid Cap Value Average Mid Value Index Performance Basis Point Differential Between High and Low 588 1, MID BLEND U.S. EQUITY INDEXES Morningstar Mid Cap Russell Mid Cap Standard & Poor's Mid Cap Wilshire Mid Cap Average Mid Blend Index Performance Basis Point Differential Between High and Low MID GROWTH U.S. EQUITY INDEXES Barra Mid Cap Growth Dow Jones Mid Cap Growth Russell Mid Cap Growth Wilshire Mid Cap Growth Average Mid Growth Index Performance Basis Point Differential Between High and Low 1,353 1,

12 Table 3. Small Cap U.S. Equity Indexes Yr. Annualized SMALL VALUE U.S. EQUITY INDEXES Barra Small Cap Value Dow Jones Small Value Russell 2000 Value Wilshire Small Value Average Small Value Index Performance Basis Point Differential Between High and Low 1,208 1, SMALL BLEND U.S. EQUITY INDEXES Morningstar Small Cap Russell Standard & Poor's Smallcap Wilshire Small Cap Average Small Blend Index Performance Basis Point Differential Between High and Low SMALL GROWTH U.S. EQUITY INDEXES Barra Small Cap Growth Dow Jones Small Growth Russell 2000 Growth Wilshire Small Growth Average Small Growth Index Performance Basis Point Differential Between High and Low 2,353 1, ,415 12

13 Table 4. Difference in 3-Year Annualized Between Best and Worst Performing U.S. Equity Index (in basis points) Value Blend Growth Large Mid Small ,415 13

14 Table Results Morningstar Equity Style Box Net Assets at the END of 2002 Net Assets ($ million) $ Amount of Actively Assets at the END of 2002 that Beat Percent of Actively Assets Beating # Actively US Equity Funds at the END of 2002 Number of Funds Number of Actively Funds at the END of 2002 that Beat Percent of Actively Funds Beating Large Value 299, , % % Large Blend 454, , % % Large Growth 397, , % % Mid Value 50,570 18, % % Mid Blend 44,728 22, % % Mid Growth 117,442 48, % % Small Value 20,449 9, % % Small Blend 64,450 53, % % Small Growth 60,691 40, % % TOTAL 1,510, , % 4,075 1, % (Redundant portfolios included) 14

15 Table Results Morningstar Equity Style Box Net Assets at the END of 2003 Net Assets ($ million) $ Amount of Actively Assets at the END of 2003 that Beat Percent of Actively Assets Beating # Actively US Equity Funds at the END of 2003 Number of Funds Number of Actively Funds at the END of 2003 that Beat Percent of Actively Funds Beating Large Value 470, , % % Large Blend 458, , % % Large Growth 623, , % 1, % Mid Value 66,697 10, % % Mid Blend 58,523 19, % % Mid Growth 184,338 66, % % Small Value 25,203 7, % % Small Blend 74,413 20, % % Small Growth 98,676 40, % % TOTAL 2,060, , % 4,921 1, % (Redundant portfolios included) 15

16 Table Results Morningstar Equity Style Box Net Assets at the END of 2004 Net Assets ($ million) $ Amount of Actively Assets at the END of 2004 that Beat Percent of Actively Assets Beating # Actively US Equity Funds at the END of 2004 Number of Funds Number of Actively Funds at the END of 2004 that Beat Percent of Actively Funds Beating Large Value 557, , % % Large Blend 552, , % % Large Growth 686, , % 1, % Mid Value 93,073 52, % % Mid Blend 71,492 32, % % Mid Growth 240, , % % Small Value 52,101 20, % % Small Blend 60,882 25, % % Small Growth 79,638 25, % % TOTAL 2,394,245 1,218, % 5,390 2, % (Redundant portfolios included) 16

17 Table 8. Three Year Average ( ) Morningstar Equity Style Box Percent of Actively ASSETS Beating Average Percent of Actively FUNDS Beating Average Differential Between Assets and Funds (in percentage points) Large Value 45.0% 34.4% 10.7% Large Blend 40.3% 36.3% 3.9% Large Growth 64.6% 51.8% 12.9% Mid Value 36.5% 37.0% -0.5% Mid Blend 42.9% 35.9% 7.0% Mid Growth 41.0% 34.4% 6.6% Small Value 39.4% 42.8% -3.4% Small Blend 50.4% 50.1% 0.3% Small Growth 46.6% 37.2% 9.4% Average for Value 40.3% 38.0% 2.2% Average for Blend 44.5% 40.8% 3.7% Average for Growth 50.7% 41.1% 9.6% Average for Large Caps 50.0% 40.8% 9.2% Average for Mid Caps 40.1% 35.8% 4.4% Average for Small Caps 45.4% 43.4% 2.1% Overall Average 49.2% 40.9% 8.2% 17

18 Figure 1. Outperforming Assets vs. Outperforming Number of Funds ( ) 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Large Value Large Blend Large Growth Mid Value Mid Blend Mid Growth Small Value Small Blend Small Growth Value Ave. Blend Ave. Growth Ave. Large Ave. Mid Ave. Small Ave. Overall Average Percent of Actively ASSETS Beating Average Percent of Actively FUNDS Beating Average 18

19 REFERENCES Arnott, Robert D, Andrew L. Berkin, and Jia Ye. How Well Have Investors Been Served in the 1980 s and 1990 s? Journal of Portfolio Management, Summer 2000, pp Brinson, Gary P., L. Randolph Hood, and Gilbert L. Beebower Determinants of Portfolio Performance. Financial Analysts Journal, vol. 51, no. 1 (January/February): Carhart, Mark. On Persistence in Mutual Fund Performance. Journal of Finance, 1997, pp Davis, James L Mutual Fund Performance and Manager Style. Financial Analysts Journal, vol. 57, no. 1 (January/February 2001): Elton, Edwin, Martin Gruber, Christopher Blake. The Persistence of Risk-Adjusted Mutual Fund Performance. Journal of Business, 1996, pp Etzioni, Ethan S. Indexing Can Be Beat. Journal of Portfolio Management, Fall 1992, pp Grinold, Richard C. and Ronald N. Kahn. The Historical Record of Active Management. Journal of Investment Consulting, 1999, vol. 2, no. 1. Gruber, Martin J. Another Puzzle: The Growth in Actively Mutual Funds. Journal of Finance, 1996, pp MacKinley, A. Craig. Asset Allocation and Stock Selection: On the Importance of Active Strategies. Journal of Investment Consulting, 1998, vol. 1, no. 1. Malkiel, Burton G. s From Investing in Equity Mutual Funds Journal of Finance, 1995, pp Pastor, Lubos and Robert Stambaugh. Investing in Equity Mutual Funds. Journal of Financial Economics, 2002, pp Sorensen, Eric H., Keith L. Miller, and Vele Samak Allocating between Active and Passive Management Financial Analysts Journal, vol. 54, no. 5 (September/October): Timbers, Stephen. The Case For Active or Passive Investment Management. Journal of Financial Planning, February 1997, pp Wermers, Russ. Mutual Fund Performance: An Empirical Decomposition Into Stock-Picking Talent, Style, Transactions Costs, and Expenses. Journal of Finance, 2000, pp

Performance Attribution: Are Sector Fund Managers Superior Stock Selectors?

Performance Attribution: Are Sector Fund Managers Superior Stock Selectors? Nicholas Scala December 2010 Abstract: Do equity sector fund managers outperform diversified equity fund managers? This paper