PII: S1057-0810(97)90004-4 FINANCIAL SERVICES REVIEW, 6(4): 257-269 Copyright 0 1997 by JAI Press Inc. MN: 1057-0810 All rights of reproduction in any form reserved. Performance of Mutual Funds Before and After Closing to New Investors Herman Mar&cyan I&ton0 Liano This study examines the decision to close mutual funds to new investors due to the growth of the funds’ assets. The evidence indicates that funds perform better three years prior to closing to new investors than they do afterwards. Furthermore, the evi- dence indicates that the closed funds outperform the control port$olios of funds with similar investment objectives and asset size during the one- and three-year periods prior to closing. However, there is no significant difference in the pe$ormance of closed finds and their matched control porcolios during the one- and three-year peri- ods after closing. Although the primary reason given for closing the finds is the desire to maintain performance in the face of growing assets, the strategy does not appear successful in accomplishing this objective. I. INTRODUCTION Since Treynor (1965), Sharpe (1966), and Jensen (1968), numerous studies have been con- ducted to examine the persistence of mutual fund performance. The results remain ambig- uous. Lehmann and Modest (1987), Grinblatt and Titman (1992), Hendricks, Patek and Zeckhauser (1993), Goetzmann and Ibbotson (1994) Brown and Goetzmann (1995) Elton, Gruber, and Blake (1996), and Gruber (1996) conclude that past performance is a good indicator of future performance. To the contrary, Jensen (1968) Malkiel (1995), Kahn and Rudd (1995), Cat-hart (1997), and Phelps and Detzel(1997) find no evidence of persistence in mutual fund performance. An important issue regarding the performance of mutual funds remains unexamined. The issue concerns the customary practice in the mutual fund industry to close funds to new investors for various reasons, some financial, and some organizational. The decision to close to new investors may be permanent or temporary, depending on the reason for the original decision. Frequently, the decision to close a fund to new investors is related to the Herman Manakyan l Western Kentucky University, Department of Accounting and Finance, 1 Big Red Way, Bowling Green, KY 42101; e-mail: Herman.Manakyan@wku.edu. Kartono Liano l Mississippi State University, Department of Finance and Economics, P.O. Box 9580, Mississippi State, MS 39762-9580; e-mail: kliano@COBILAN.MsState.Edu. 258 FINANCIAL SERVICES REVIEW 6(4) 1997 fund’s asset size. The most common explanation given for closing funds to new investors is the difficulty in effectively managing a large portfolio to earn superior returns for the investors. Presumably, once a portfolio reaches a critical threshold of size, it becomes “too bulky” to manage effectively, and it is difficult to find attractive investments to which new funds can be diverted without sacrificing performance. If this explanation is valid, mutual funds that close to new investors should be able to at the very least maintain performance after the decision to close. However, the views on closing mutual funds to new investors are far from unanimous in the mutual fund industry. While some investment companies follow this practice regu- larly, there seems to be no consistency in how large a fund is allowed to grow prior to clos- ing. Other investment companies eschew this practice completely, allowing the funds to grow without limit, while continuing to post enviable investment results for their clients. An interesting example is Fidelity Magellan Fund, the largest of all mutual funds, which was allowed to grow in excess of $60 billion in assets and for many years posted impres- sive rates of return far exceeding the returns generated by smaller, more nimble funds. Magellan was finally closed to new investors effective September 30, 1997. The existing literature on mutual funds provides no empirical evidence on the impact of this widespread practice on mutual fund performance. The purpose of this study is to empirically examine whether the decision to close mutual funds to new investors in order to limit the growth of assets is justified by the subsequent performance of the funds relative to both their own historical performance and the performance of their peers. II. METHODOLOGY AND DATA In order to examine the impact of closing mutual funds to new investors, it was first neces- sary to identify a sample of mutual funds that are closed to new investors, along with the date of and the stated reason for closure. The January 1996 version of the MORNINGSTAR On Disc data base was screened to identify all funds with purchase constraint codes of L (closed to all investors) and C (closed to new investors). The resulting 128 funds were then contacted to inquire about the date on which they stopped accepting new investors and the reason given for closing. In order to be able to examine the performance of the funds for a period of at least one year on either side of the closing date, it was necessary to limit the study to funds that closed between 1978 and 1994. Of the 63 funds that closed to new investors during this time period, 38 indicated that the decision to close was related to the fund’s asset size, while 25 funds expressed other motives for closing, the most frequent of which was a change in the fund’s fee structure. The sample was reduced further due to some of the funds having closed and re-opened to new investors within the analysis win- dow, resulting in overlapping periods, and due to lack of data. The final sample consisted of 27 funds that allowed an analysis of a one-year window around the closing date, and 12 of those funds had sufficient data for the analysis of a three-year window around the clos- ing date. The funds included in the analysis are identified in Table 1, along with their investment objectives and month of closing to new investors. The same source was used to form matching control portfolios for each fund in the sample. The control portfolios consisted of the five funds within the same investment objective with net assets closest to the sample fund, which were open to new investors and Performance of Mutual Funds 259 had returns data available for the relevant analysis window. The monthly returns of the sample funds and the matching control portfolios for the one (three) year(s) surrounding the date of closing were then extracted from MORNZNGSTAR. The three-month T-Bill returns were extracted from the Federal Reserve BuEletin, and the monthly returns of the S&P 500 Index were extracted from the Security Price Index Record. The impact of the decision to close the funds to new investors was examined by com- paring mean raw returns, as well as widely used risk adjusted portfolio performance mea- sures for the l-year and 3-year periods prior to and following the closing of the fund to new investors. Sharpe’s index was used to measure performance adjusted by total risk, and Treynor’s index was used to measure performance adjusted by systematic risk. Jensen’s alpha was included as an additional indexed performance measure. Fabozzi, Francis, and Lee (1980) demonstrate that the Jensen performance measure is robust when returns are measured monthly. In addition, Fabozzi and Francis (1979) conclude that Jensen’s alpha is not influenced by bull and bear markets. However, Barber (1994) finds that stock mutual funds’ future returns are not related to the historical beta of the fund. Since the sample includes fixed income and foreign funds, the 01, l3, and Treynor’s index for these funds should be interpreted with caution, as the S&P 500 index was used as the market return proxy for all sample funds for consistency. Symbol TABLE 1 Sample of Mutual Funds Closed to New Investments Fund Name Objective Closing Month ACINX ACRNX ARGFX BABEX BARIX CPGRX CPSFX IDFDX JAVLX JAVTX KRFBX LOMCX MMHYX MNSCX MONTX MPSCX MSIQX NEFGX PJIGX POPAX POPCX SEQUX SKSEX SSRSC STCSX VWNDX Van Kampen Am Cap Ltd Mat B Acorn International Acorn Ariel Growth Babson Enterprise Baird Adjustable Rate Income Chesapeake Growth Comstock Partners Strategy 0 Idex 3 Janus Twenty Janus Venture Kemper Retirement II CGM Capital Development MFS Municipal High-Income A Montgomery Small Cap Monetta MAS Small Cap Value Morgan Stanley Instl Intl Equity New England Growth A Piper Jaffray Instl Govt Pimco Adv Opportunity A Pimco Adv Opportunity C Sequoia Skyline Special Equities State St Research Sm Cap GrC Strong Common Stock Vanguard Windsor Adi. Rate Mtg. 5193 Foreign _ 2194 Small Company 7190 Small Company 4190 Small Company II92 Adj. Rate Mtg. 12194 Growth 12194 Mult-Asst Glbl 7192 Growth 6190 Growth l/93 Small Company 9191 Balanced 3192 Growth 11186 Muni Nat’1 6185 Small Company 3192 Small Company 3193 Small Company 9194 Foreign 6193 Growth l/92 Gvt Mortgage 6194 Aggr Growth 12192 Aggr Growth 12192 Growth 12182 Small Company 12192 Small Company 1 l/94 Small Company 3193 Growth-Income 5/85 260 FINANCIAL SERVICES REVIEW 6(4) 1997 In addition to the raw return (unadjusted for risk), each of the three risk-adjusted per- formance measures were employed to conduct three types of comparison for a total of twelve comparisons: (1) One- and three-year performance of each closed fund relative to itself (pre- vs. post-closing). (2) One- and three-year pre-closing performance of each closed fund relative to the control portfolio. (3) One- and three-year post-closing performance of each closed fund relative to the control portfolio. Due to the small sample size, the non-parametric Wilcoxon rank sum test, which is more powerful for small samples of unknown distribution (Gibbons, 1985, p. 193), is used for these pairwise comparisons. III. RESULTS Table 2 presents the performance measures (the raw return, Sharpe’s index, Jensen’s alpha, and Treynor’s index), as well as beta, and R* values for the 27 funds during the 12 months before and after the closing of a fund to new investors. As expected, the equity funds in the sample have beta values near one, and fixed income and foreign funds have lower betas. The mean returns during the 12 months before closing are higher than the mean returns during the 12 months after closing for 16 of the 27 funds. The average monthly raw return 12 months prior to closing is 1.63% and is significantly different from zero at the I % level. In comparison, the average monthly raw return 12 months after closing is 1.17% and is sig- nificantly different from zero at the 1% level. On the average, mutual funds generate an additional annual return of 5.64% before closing to new investors than after closing the funds ([l + (0.016268-0.011684)]‘* - 1). The non-parametric Wilcoxon rank sum test reveals that the raw return during the year before closing is not significantly different from the raw return during the year after closing of the funds. Furthermore, the average Jensen’s alpha before closing is positive and signif- icantly different from zero at the 1% level, implying that before closing to new investors, the funds beat the market portfolio, as measured by the S&P 500 index. However, the Jensen’s alpha after closing the funds is not significantly different from zero, suggesting that the funds fail to beat the market portfolio after closing to new investors. Based on raw returns, the Sharpe’s index, Jensen’s alpha, and Treynor’s index, 16, 11, 19, and 16 of the 27 funds, respectively, performed better during the 12 months preceding the closing of the fund than the 12 subsequent months. However, the non-parametric Wil- coxon rank sum test indicates that the performance of the sample funds in the 12 months preceding their closing is not statistically superior at the usual significance levels relative to their performance after closing to new investors. The results are similar when comparing performance in the 36 months preceding and following the closing of the funds for the 12 funds for which the three-year comparison was possible. The mean returns are higher during the three years before closing than the three years after closing for 10 of the 12 funds. The average monthly raw returns during the three T A B L E 2 T he I - Y ea r Pe rf or m an ce Su m m ar y B ef or e an d A ft er C lo si ng B ef or e A ft er i R aw Sh ar pe Je ns en T re yn or B et a R 2 R aw Sh ar pe Je ns en T re yn or B et a R 2 4 A C FT X 0. 20 75 -0 .2 49 4 -0 .0 48 6 1. 52 44 -0 .0 38 0 0. 11 36 -0 .0 28 3 -0 .8 04 1 -0 .2 87 2* ** -2 .5 10 6 0. 12 07 ** * 0. 54 72 A C IN X 3. 71 42 1. 21 06 3. 01 44 ** * 4. 20 63 0. 82 25 0. 29 I 2 -1 a 43 3 -0 .5 25 3 -1 .4 41 5 -2 .4 93 8 0. 57 84 ** 0. 39 52 2 A C R N X I . 06 50 0. 09 69 0. 05 36 0. 48 73 0. 82 39 ** * 0. 81 70 0. 60 50 0. 00 93 -0 .3 09 9 0. 04 86 1. 30 20 ** * 0. 90 74 A R G FX 0. 94 58 0. 07 25 -0 .2 57 3 0. 29 78 0. 89 97 ** * 0. 90 31 0. 96 83 0. 05 10 -0 .3 75 9 0. 29 78 1. 27 58 ** * 0. 78 44 [ B A B E X 3. 10 75 0. 62 34 I . 34 57 3. 25 1 8 0. 81 47 ** * 0. 76 52 1. 47 67 .0 38 66 0. 89 85 I . 32 39 0. 89 94 ** 0. 34 81 B A R IX -0 .4 32 5 -0 .5 55 I -0 .8 59 7 4. 13 12 -0 .1 87 8 0. I 63 6 I . 37 08 0. 52 15 0. 31 61 3. 10 49 0. 29 07 0. 06 18 5 C PG R X I . 26 33 0. 16 39 I . 48 94 0. 72 17 I . 27 48 ** 0. 48 1 2 2. 39 50 0. 31 00 - 1 .3 56 5 I. 18 41 I . 62 73 0. 14 91 b C PS FX 0. 68 50 0. 23 79 0. 26 I 2 3. 01 86 0. 10 36 0. I O 76 0. 72 67 0. 3 I3 2 0. 54 97 -I . 30 26 -0 .3 61 9 0. 16 35 ID F D X 1. 67 17 0. 16 88 0. 50 03 0. 85 14 1. 18 06 ** * 0. 80 99 0. 83 42 0. 04 64 0. 44 97 0. 23 52 1. 18 08 ** * 0. 92 43 JA V L X 0. 20 50 -0 .0 28 I -0 .1 82 6 -0 .0 85 8 1. 01 03 ** 0. 49 56 0. 54 67 0. 09 20 -0 .3 41 8 0. 24 96 1. 17 10 ** 0. 47 85 JA V T X 3. 11 92 0. 66 16 1. 47 71 ** 2. 90 75 0. 89 17 ** * 0. 75 25 0. 59 00 0. 08 02 0. 00 42 0. 37 87 0. 70 48 ** * 0. 74 47 K R FB X I . 90 67 0. 40 76 I. 01 91 2. 05 24 0. 72 15 ** * 0. 74 53 I. 14 75 0. 51 06 0. 57 I 6 I . 96 96 0. 44 47 0. 22 62 L O M C X 3. 54 75 0. 44 16 I. 11 12 2. 69 95 I. 1 14 8* ** 0. 72 69 0. 52 1 7 0. 00 23 I. 10 71 0. 02 22 I . 37 97 ** * 0. 81 42 M M H Y X I . 44 33 0. 40 33 0. 68 57 -9 3. 56 59 -0 .0 07 2 0. 00 03 1. 31 67 0. 68 00 0. 66 63 19 .7 88 7 0. 03 69 0. 10 15 7 M N SC X 4. 29 25 0. 61 65 3. 07 78 ** 3. 13 68 1. 23 27 ** * 0. 72 90 0. 94 75 0. 13 82 -0 .1 86 0 0. 53 65 I . 25 97 0. 22 43 M O N T X -0 .5 oo o -0 .2 44 7 -0 .9 91 I -1 .2 48 1 0. 62 45 0. 15 70 0. 44 17 0. 05 50 0. 51 I 4 0. 18 58 0. 97 47 ** * 0. 57 19 M PS C X I . 07 00 0. 23 74 0. 81 1 9 0. 96 84 0. 79 30 ** 0. 43 84 I . 46 67 0. 29 92 -0 .2 76 7 1. 18 99 0. 83 88 0. 29 88 M SI Q X 0. 86 58 0. 15 73 0. 76 24 -I .6 62 3 -0 .3 64 1 0. 03 77 2. 38 17 0. 48 18 2. 54 35 ** 1. 74 39 I . 20 38 ** 0. 48 62 (c on ti nu ed ) T A B L E 2 (C on ti nu ed ) B ef or e A ft er R aw Sh ar pe Je ns en T re yn or B et a R 2 R aw Sh ar pe Je ns en T re yn or B et a R 2 N E FG X 3. 97 08 0. 59 49 1. 53 47 ** 2. 84 15 1. 23 62 ** * 0. 90 62 0. 04 42 -0 .0 88 5 -0 .5 95 7 -0 .2 22 0 1. 08 92 ** * 0. 64 74 PJ IG X - 1 .7 90 0 -0 .4 64 5 -2 .0 00 4 -4 .2 14 5 0. 48 99 0. 06 87 0. 97 25 0. 19 38 -0 .4 86 6 0. 67 03 0. 78 03 ** * 0. 52 32 PO PA X 3. 06 75 0. 43 69 1. 70 22 2. 42 52 1. 14 21 ** 0. 46 49 2. 78 08 0. 50 24 2. 03 72 1. 69 84 1. 48 75 0. 25 58 PO PC X 3. 00 50 0. 42 64 1. 63 53 2. 36 08 1. 14 68 ** 0. 46 72 2. 72 00 0. 49 1 8 1. 97 50 1. 65 27 1. 49 18 0. 25 89 SE Q U X 2. 06 50 0. 36 98 1. 14 24 ** * 2. 20 46 0. 51 26 ** * 0. 87 18 2. 05 00 0. 64 80 1. 02 75 3. 00 77 0. 43 61 ** 0. 39 42 SK SE X 3. 23 83 0. 66 87 2. 43 39 5. 42 48 0. 54 2 1 0. 21 74 1. 74 92 0. 70 33 1. 18 47 ** 1. 58 55 0. 94 27 ** * 0. 57 58 SS R SC -1 .1 30 8 -0 .3 1 83 1. 27 46 - 1 .6 60 5 0. 87 84 0. 29 56 1. 31 92 0. 16 17 - 1 .7 76 8 0. 63 69 1. 33 10 0. 14 61 ST C SX 1. 27 75 0. 29 44 0. 63 92 0. 94 26 1. 05 88 ** 0. 39 69 0. 91 25 0. 21 56 0. 98 26 0. 66 8 1 0. 97 57 ** * 0. 67 82 V W N D X 2. 04 17 0. 32 60 1. 03 05 ** 1. 50 94 0. 83 35 ** * 0. 86 94 2. 33 41 0. 55 79 0. 30 80 2. 07 60 0. 83 93 ** * 0. 91 26 A ve ra ge 1. 62 68 ** * 0. 25 02 ** * 0. 74 50 ** * -2 .0 17 5 0. 72 41 ** * 0. 48 49 ** * 1. 16 84 ** * 0. 22 35 ** * 0. 28 51 1. 39 73 0. 90 00 ** * 0. 46 42 ** * V ar ia nc e 2. 57 7 1 0. 15 25 1. 46 80 33 9. 06 08 0. 21 23 0. 09 45 0. 80 00 0. 11 98 1. 06 65 15 .2 97 7 0. 24 1 1 0. 07 43 N o te s: W ilc o xo n Z -s ta ti st ic fo r R aw R et u rn s: 1. 29 75 (0 .1 94 5) W ilc o xo n Z -s ta ti st ic fo r Je n se n ’s A lp h a: I .7 81 9 (0 .0 74 8) W ilc o xo n Z -s ta ti st ic fo r S h ar p e’ s In d ex 0. 25 95 (0 .7 95 2) W ilc o xo n Z -s ta ti st ic fo r T re yn o r’ s In d ex : 1. 46 19 (0 .1 43 8) p -v al u es ar e in p ar en th es es ** *S ig n ifi ca n tly d if fe re n t fr o m ze ro at t h e 1% le ve l ** S ig n iti ca n tly d if fe re n t fr o m ze ro at t h e 5% le ve l Perfornmnce of Mutual Funds 263 years before and the three years after closing are 1.63% and 1.09%, respectively, and are significantly different from zero at the 1% level.’ On the average, mutual funds earn an additional annual return of 6.57% per year for a three-year period before closing to new investors than after closing the funds. The non- parametric Wilcoxon rank sum test reveals that the raw return during the three years before closing is significantly different from the raw return during the three years after closing at the 5% level. The Jensen’s alpha before and after closing is positive and significantly dif- ferent from zero at the 1% and 5% level, respectively, indicating that the funds beat the market portfolio before and after closing to new investors. Based on the raw returns, Jensen’s alpha and Treynor’s index, 10, eight, and nine of the 12 funds, respectively, per- formed better prior to closing. The Wilcoxon test indicates that the difference in the perfor- mance of the funds before and after closing is statistically significant at the 5% level. The comparison of Sharpe’s index is inconclusive, as seven of the 12 funds perform better prior to closing, but the test statistic is not significant. Based on the comparison of performance measures before and after the closing of a fund to new investors, there is no evidence to indicate that closing of the fund resulted in improved performance. To the contrary, fund performance seems to have deteriorated after closing to new investors. These findings are consistent with the observations of Damato and Jereski (1997) who compare the one-year return of 13 equity funds with more than $500 million in assets that are closed to new investors since 1989 with the S&P 500 index, and conclude that closing a fund does not guarantee performance. To further examine the impact of the practice of closing funds to new investors as the assets of the fund grow, the performance of each fund in the sample was compared to a matched control portfolio of funds. The control portfolios consisted of the five funds within the same investment objective with asset size closest to the sample fund and having sufficient monthly returns data to allow a one- or three-year comparison. In Table 3, the performance measures of the 27 funds in the sample during the 12- month period prior to closing of the funds are compared to the performance measures of the matching control portfolios during the same time frame. Using the raw return, Sharpe’s index, Jensen’s alpha, and Treynor’s index, the results indicate that 19 of the 27 funds out- performed their matched portfolios during the year preceding the closing of the funds. However, the Wilcoxon test suggests that the performance of the sample funds in the 12 months preceding their closing is not statistically different at the usual significance levels from the performance of control portfolios. When the analysis is extended to the 36 months before the closing of the funds, the results are stronger. Of the 12 funds in this sample, 11 outperformed their control portfolios based on the raw return, Sharpe’s index, and Treynor’s index and 10 outperformed their control portfolios based on the Jensen’s alpha. The Wilcoxon rank sum test is significant at the 1% level for the Treynor’s index and at the 5% level for the other performance mea- sures, providing strong evidence that the funds in the sample were superior in performance to funds with similar investment objectives and size. In contrast, when the sample funds are compared to their control portfolios after clos- ing to new investors, their performance does not compare as favorably to the control port- folios. Table 4 indicates that during the 12 months following their closing to new investments, only 11 of the 27 sample funds outperformed their matched control portfolios, measured by the Sharpe’s index. The Jensen’s alpha measure indicates 12 of the 27 sample funds outperformed their matched portfolios. In addition, the raw return and the Treynor’s T A B L E 3 T he l- Y ea r Pe rf or m an ce C om pa ri so n W ith M at ch ed Po rt fo lio s B ef or e C lo si ng R aw A C F T X Sa m pl e Fu nd s M at ch ed P or rf ol io s Sh L lr pe Je ns en T re yn or B et a R 2 R aw Sh ar pe Je n se n T re yn or B et a R 2 0. 20 75 -0 .2 49 4 -0 .0 48 6 1. 52 44 -0 .0 38 0 0. 11 36 0. 44 67 0. 56 7 1 0. 19 92 ** -2 .4 81 1 -0 .0 73 0 0. 22 38 A C IN X 3. 71 42 A C R N X 1. 06 50 A R G F X 0. 94 58 B A B E X 3. 10 75 B A R IX -0 .4 32 5 C P G R X 1. 26 33 C P S F X 0. 68 50 ID F D X 1. 67 17 JA V L X 0. 20 50 JA V T X 3. 11 92 K R F B X 1. 90 67 L O M C X 3. 54 75 M M H Y X 1. 44 33 M N S C X 4. 29 25 M O N T X -0 .5 00 0 M P S C X 1. 07 00 M S IQ X 0. 86 58 N E F G X 3. 97 08 P JI G X 1. 79 00 1. 21 06 0. 09 69 0. 07 25 0. 62 34 -0 .5 55 1 0. 16 39 0. 23 79 0. 16 88 -0 .0 28 1 0. 66 16 0. 40 76 0. 44 16 0. 40 33 0. 61 65 -0 .2 44 7 0. 23 74 0. 15 73 0. 59 49 -0 .4 64 5 3. 01 44 ** * 4. 20 63 0. 82 25 0. 29 12 3. 32 33 0. 05 36 0. 48 73 0. 82 39 ** * 0. 81 70 1. 72 25 -0 .2 57 3 0. 29 78 0. 89 97 ** * 0. 90 31 1. 34 17 1. 34 57 3. 25 18 0. 81 47 ** * 0. 76 52 3. 06 67 -0 .8 59 7 4. 13 12 -0 .1 87 8 0. 16 36 0. 02 42 1. 48 94 0. 72 17 1. 27 48 ** 0. 48 12 0. 15 17 0. 26 12 3. 01 86 0. 10 36 0. 10 76 0. 55 25 0. 50 03 0. 85 14 1. 18 06 ** * 0. 80 99 1. 33 25 -0 .1 82 6 -0 .0 85 8 1. 01 03 ** 0. 49 56 1. 22 08 1. 47 71 ** 2. 90 75 0. 89 17 ** * 0. 75 25 2. 61 25 1. 01 91 2. 05 24 0. 72 15 ** * 0. 74 53 1. 13 08 1. 11 12 2. 69 95 1. 11 48 ** * 0. 72 69 2. 46 17 0. 68 57 -9 3. 56 59 -0 .0 07 2 0. 00 03 1. 28 67 3. 07 78 ** 3. 13 68 1. 23 27 ** * 0. 72 90 2. 85 58 -0 .9 91 1 -1 .2 48 1 0. 62 45 0. 15 70 0. 40 92 0. 81 19 0. 96 84 0. 79 30 ** 0. 43 84 0. 26 00 0. 76 24 1. 66 23 -0 .3 64 1 0. 03 77 0. 21 17 1. 53 47 ** 2. 84 15 1. 23 62 ** * 0. 90 62 2. 51 75 -2 .0 04 -4 .2 14 5 0. 48 99 0. 06 87 0. 07 08 0. 78 38 2. 46 44 ** 2. 74 87 1. 11 65 0. 28 61 0. 20 02 0. 60 76 0. 99 09 1. 06 87 ** * 0. 84 30 0. 13 06 0. 00 19 0. 58 55 1. 13 37 ** * 0. 75 74 0. 54 54 1. 10 59 2. 62 08 0. 99 53 ** * 0. 89 75 -1 .5 25 1 -0 .3 02 8* ** -8 .6 97 0 0. 03 67 0. 23 12 -0 .0 70 7 0. 18 39 -0 .2 27 9 0. 84 09 ** * 0. 89 72 0. 12 18 0. 02 24 0. 56 84 0. 31 71 ** * 0. 77 24 0. 13 16 0. 21 52 0. 63 18 1. 05 40 ** * 0. 89 35 0. 49 25 0. 87 35 1. 58 59 0. 58 59 ** 0. 44 41 0. 35 87 0. 28 5 1 1. 44 91 1. 43 95 ** * 0. 89 27 0. 25 80 0. 31 03 1. 14 32 0. 61 66 ** * 0. 95 41 0. 38 38 0. 32 43 2. 04 80 0. 93 92 ** * 0. 95 77 0. 41 80 0. 42 97 7. 10 83 0. 07 32 0. 04 81 0. 50 69 1. 80 90 ** 2. 50 42 0. 97 03 ** * 0. 76 89 0. 02 67 -0 .2 70 5 0. 10 98 1. 18 11 0. 24 16 -0 .0 11 2 0. 01 59 -0 .0 40 1 1. 04 63 ** * 0. 56 84 -0 .0 18 6 0. 01 56 0. 32 78 -0 .1 49 5 0. 01 34 0. 43 65 0. 42 49 2. 01 60 1. 02 14 ** * 0. 97 03 -0 .2 53 4 -0 .1 73 8 -0 .8 89 1 0. 22 94 ** 0. 44 95 P O P A X 3. 06 75 0. 43 69 1. 70 22 2. 42 52 P O P C X 3. 00 50 0. 42 64 1. 63 53 2. 36 08 S E Q U X 2. 06 50 0. 36 98 1. 14 24 ** * 2. 20 46 S K S E X 3. 23 83 0. 66 87 2. 43 39 5. 42 48 S S R S C -1 .1 30 8 -0 .3 1 83 -1 .2 74 6 -1 .6 60 5 S T C S X 1. 27 75 0. 29 44 0. 63 92 0. 94 26 V W N D X 2. 04 17 0. 32 60 1. 03 05 ** 1. 50 94 A ve ra ge 1. 62 68 ** * 0. 25 02 ** * 0. 74 50 ** * -2 .0 17 5 V ar ia n ce 2. 57 71 0. 15 25 1. 46 80 33 9. 06 08 N ot es : W ilc ox on Z -s ta tis tic fo r R aw R et ur ns : 0. 92 56 (0 .3 54 7) W ilc ox on Z -s ta tis tic fo r S ha rp e’ s In de x 0. 72 66 (0 .4 67 5) p- va lu es ar e in p ar en th es es ** *S ig ni fi ca nt ly di ff er en t fr om z er o at t he I % l ev el ** Si gn if ic an tly di ff er en t fr om z er o at t he 5 % l ev el 1. 14 21 ** 0. 46 49 1. 04 58 0. 18 19 -0 .0 37 6 1. 14 68 ** 0. 46 72 1. 51 75 0. 22 42 0. 11 44 0. 51 26 ** * 0. 87 18 1. 65 08 0. 13 21 0. 73 87 ** 0. 54 21 0. 21 74 1. 63 92 0. 24 18 0. 36 63 0. 87 84 0. 29 56 0. 29 50 -0 .0 08 7 0. 17 89 1. 05 88 ** 0. 39 69 0. 92 58 0. 16 55 0. 25 2 1 0. 83 35 ** * 0. 86 94 1. 44 33 0. 18 76 0. 43 80 ** * 0. 72 41 ** * 0. 48 49 ** * 1. 31 54 ** * 0. 17 06 ** 0. 38 90 ** * 0. 21 23 0. 09 45 0. 95 09 0. 16 78 0. 36 04 W ilc ox on Z -s ta tis tic fo r Je ns en ’s A lp ha : 1. 18 65 ( 0. 06 93 ) W ilc ox on Z -s ta tis tic fo r T re yn or ’s In de x: 1. 41 86 (0 .1 56 0) 0. 89 00 0. 84 06 ** * 1. 03 15 1. 18 25 ** * 0. 75 67 0. 94 60 ** * 1. 28 59 1. 04 32 ** * -0 .0 32 3 1. 01 04 ** * 0. 55 56 1. 16 34 ** 0. 81 26 0. 81 19 ** * 0. 71 86 0. 79 41 ** * 6. 20 45 K 18 91 0. 59 74 % 0. 67 61 9 0. 96 69 0. 50 56 0. 57 71 i 4 0. 36 25 0. 98 89 : 0. 62 17 ** * i 0. 09 34 f? , 9 & R aw A C F T X T A B L E 4 T he 1 -Y ea r Pe rf or m an ce C om pa ri so n W ith M at ch ed Po rt fo lio s A ft er C lo si ng Sa m pl e F un ds M at ch ed P or tf ol io s Sh ar pe Je ns en T re yn or B et a R 2 R aw S h ar pe Je ns en T re yn or B et a R 2 -0 .0 28 3 -0 .8 04 1 -0 .2 87 2* ** -2 .5 10 6 0. 12 07 ** * 0. 54 72 0. 02 75 -0 .5 4 18 -0 .2 38 4 -3 .6 55 6 0. 06 76 0. 11 09 A C IN X - 1 .0 43 3 -0 .5 25 3 -1 .4 41 5 -2 .4 93 8 A C R N X 0. 60 50 0. 00 93 -0 .3 09 9 0. 04 86 A R G F X 0. 96 83 0. 05 10 -0 .3 75 9 0. 29 78 B A B E X 1. 47 67 0. 38 66 0. 89 85 1. 32 39 B A R IX 1. 37 08 0. 52 15 0. 31 61 3. 10 49 C P G R X 2. 39 50 0. 31 00 - 1 .3 56 5 1. 18 41 C P S F X 0. 72 67 0. 31 32 0. 54 97 -1 .3 02 6 ID F D X 0. 83 41 0. 04 64 0. 44 97 0. 23 52 JA V L X 0. 54 67 0. 09 20 -0 .3 41 8 0. 24 96 JA V T X 0. 59 00 0. 08 02 0. 00 42 0. 37 87 K R F B X 1. 14 75 0. 5 10 6 0. 57 16 1. 96 96 L O M C X 0. 52 17 0. 00 23 1. 10 71 0. 02 22 M M H Y X 1. 13 67 0. 68 00 0. 66 63 19 .7 88 7 M N S C X 0. 94 75 0. 13 82 -0 .1 86 0 0. 53 65 M O N T X 0. 44 17 0. 05 50 0. 51 14 0. 18 58 M P S C X 1. 46 67 0. 29 92 -0 .2 76 7 1. 18 99 M S IQ X 2. 38 17 0. 48 18 2. 54 35 ** 1. 74 39 0. 57 84 ** 0. 39 52 1. 30 20 ** * 0. 90 74 1. 27 58 ** * 0. 78 44 0. 89 94 ** 0. 34 8 1 0. 29 07 0. 06 18 1. 62 73 0. 14 91 -0 .3 61 9 0. 16 35 1. 18 08 ** * 0. 92 43 1. 17 10 ** 0. 47 85 0. 70 48 ** * 0. 74 47 0. 44 47 0. 22 62 1. 37 97 ** * 0. 81 42 0. 03 69 0. 01 57 1. 25 97 0. 22 43 0. 97 47 ** * 0. 57 19 0. 83 88 0. 29 88 1. 20 38 ** 0. 48 62 -0 .8 52 5 -0 .3 80 1 1. 25 03 - 1 .6 00 9 0. 78 17 ** * 0. 50 40 1. 67 33 0. 12 57 0. 63 88 0. 65 81 1. 71 94 ** * 0. 91 65 1. 54 92 0. 11 96 0. 09 78 0. 65 96 1. 45 66 ** * 0. 88 04 0. 59 83 0. 08 68 -0 .0 20 1 0. 30 52 1. 02 34 0. 32 86 0. 54 33 0. 32 58 0. 04 25 4. 63 89 0. 01 62 0. 01 17 2. 25 33 0. 82 86 -0 .1 05 9 1. 90 47 0. 93 72 ** 0. 41 12 0. 91 58 0. 58 19 0. 63 10 4. 82 65 0. 13 69 0. 04 11 0. 64 67 0. 01 39 0. 27 88 0. 06 97 1. 29 51 ** * 0. 94 91 1. 11 08 0. 41 70 0. 37 62 0. 96 56 0. 88 70 ** * 0. 65 74 0. 63 83 0. 06 82 -0 .0 13 2 0. 35 78 0. 88 11 ** * 0. 60 34 1. 18 58 0. 68 25 0. 47 89 ** I . 43 70 0. 63 61 ** * 0. 76 01 -0 .1 95 8 -0 .0 79 2 0. 09 17 -0 .6 88 3 0. 99 79 ** * 0. 97 86 1. 26 25 0. 38 65 0. 24 94 2. 74 28 0. 24 65 0. 25 64 0. 77 33 0. 13 92 -0 .1 12 6 0. 55 87 0. 89 79 0. 20 92 0. 98 42 0. 19 05 1. 12 20 0. 61 55 1. 17 56 ** * 0. 62 49 2. 64 67 0. 62 95 0. 91 42 2. 61 90 0. 83 17 0. 27 23 1. 78 50 0. 30 92 2. 09 80 ** 0. 93 14 1. 61 34 ** * 0. 70 23 N E FG X 0. 04 42 -0 .0 88 5 -0 .5 95 7 -0 .2 22 0 PJ IG X 0. 97 25 0. 19 38 -0 .4 86 6 0. 67 03 PO PA X 2. 78 08 0. 50 24 2. 03 72 1. 69 84 PO PC X 2. 72 00 0. 49 18 1. 97 50 1. 65 27 SE Q U X 2. 05 00 0. 64 80 1. 02 75 3. 00 77 SK SE X 1. 74 92 0. 70 33 1. 18 47 ** 1. 58 55 SS R SC 1. 31 92 0. 16 17 - 1 .7 76 8 0. 63 69 ST C SX 0. 91 25 0. 21 56 0. 98 26 0. 66 8 1 V W N D X 2. 33 41 0. 55 79 0. 30 80 2. 07 60 A ve ra ge 1. 16 84 ** * 0. 22 35 ** * 0. 28 51 1. 39 73 * V ar ia nc e 0. 80 00 0. 11 98 1. 06 65 15 .2 97 7 N ot es : W ilc ox on Z -s ta tis tic fo r R aw R et ur ns : -0 . I S. 57 (0 .8 76 3) W ilc ox on Z -s ta tis tic fo r S ha rp e’ s In de x -0 .5 53 6 (0 .5 79 9) p- va lu es ar e in p ar en th es es ** *S ig ni fi ca nt ly di ff er en t fr om z er o at t he 1 % l ev el ** Si gn if ic an tly di ff er en t fr om z er o at t he 5 % l ev el 1. 08 92 ** * 0. 64 74 1. 10 17 0. 32 78 0. 47 55 0. 78 03 ** * 0. 52 32 0. 81 83 0. 37 97 0. 02 89 1. 48 75 0. 25 58 1. 38 25 0. 32 68 0. 74 24 1. 49 18 0. 25 89 1. 13 25 0. 26 07 0. 42 1 1 0. 43 61 ** 0. 39 42 1. 71 00 0. 28 63 0. 27 57 0. 94 27 ** * 0. 57 58 1. 19 83 0. 26 74 0. 54 01 1. 33 10 0. 14 61 3. 28 33 0. 82 30 0. 84 48 0. 97 57 ** * 0. 67 82 0. 97 50 0. 23 04 1. 07 85 ** 0. 83 93 ** * 0. 91 26 2. 52 25 0. 60 84 0. 42 04 ** 0. 90 00 ** * 0. 46 42 ** * 1. 17 30 ** * 0. 27 46 ** * 0. 37 43 ** * 0. 24 11 0. 07 43 0. 75 91 0. 10 00 0. 34 89 W ilc ox on Z -s ta tis tic fo r Je ns en ’s A lp ha : -0 .5 01 7 (0 .6 15 9) W ilc ox on Z -s ta tis tic fo r T re yn or ’s In de x: -0 .6 40 1 (0 .5 22 1) 0. 77 91 1. 04 71 ** * I . 40 38 0. 26 28 ** 0. 96 18 1. 17 27 ** 0. 63 19 1. 38 95 ** 0. 90 99 1. 06 80 ** * 0. 76 87 1. 22 77 ** 2. 81 88 0. 99 76 0. 66 50 1. 07 43 ** * 2. 18 47 0. 88 37 ** * 1. 05 44 ** * 0. 91 57 ** * 2. 82 07 0. 20 12 0. 71 89 P 0. 47 14 s 0. 33 76 0. 49 77 0. 83 15 [ 4 0. 35 39 0. 19 32 0. 78 17 2 0. 97 98 h 0. 53 27 ** * G 0. 08 94 f 8 268 FINANCIAL SERVICES REVIEW 6(4) 1997 index are higher for 13 of the 27 sample funds over the matched portfolios. However, the Wilcoxon test indicates no statistical difference in performance between the sample funds and the control portfolios. When the analysis is extended to three years around the closing date, the results are similar. Based on the Treynor’s index only six of 12 funds outperformed their control port- folios after closing to new investors. Using the raw return, Sharpe’s index, and Jensen’s alpha, seven of 12 funds outperformed their peers in the 36 months following the decision to close to new investors. The test statistic indicates no significant differences in perfor- mance between the sample funds and the control portfolios. Thus, while the sample funds significantly outperformed their peers prior to closing, the performance does not seem to persist after closing to new investors. Consequently, the empirical evidence does not sup- port the wisdom of closing funds to new investors for reasons related to asset size. IV. IMPLICATIONS OF THE STUDY Mutual funds are the most popular investment outlets for individual investors. As such, it is important to develop a thorough understanding of factors that influence the performance of mutual funds. While many facets of mutual fund performance have been examined in the finance literature, the common practice of closing funds to new investors as funds grow has not received any attention in the literature. Fund managers often justify the decision to close funds to new investors based on the desire to maintain superior performance. The premise underlying this line of reasoning is that due to limited suitable investment opportunities, substantial amounts of assets or con- tinued substantial cash inflows make it difficult to maintain strong performance. However, the empirical evidence presented here contradicts the premise that superior performance can be maintained by restricting access to the funds, or by capping the asset size. The evi- dence indicates that based on the raw return, Sharpe’s index, Jensen’s alpha, and Treynor’s index, funds performance declines after closing to new investors. The conclusion is the same when the sample funds are compared to control portfolios matched in size and invest- ment objective. Thus, maintaining performance does not seem to be an appropriate reason for closing funds to new investors. The practice of closing funds to new investors continues to be popular. While the ini- tial sample of this study was the 128 funds closed to investors as of January 1996, MORN- INGSTAR data indicates there were 141 funds closed to new investors as of January 1998. However, the actual number of funds closed during the interim is significantly higher, as many funds close to new investors for short periods, and subsequently re-open. Thus, there are a number of funds that were closed in January 1996, that are now open to the public, as well as a large number of funds that have closed to new investors since that time. Of the 27 funds that were used in the one-year comparisons, 17 continue to be closed to new invest- ments as of January 1998, and three are no longer available in MORNZNGSTAR. Similarly, of the 12 funds with data available to facilitate the three-year analysis, eight remain closed to new investments and one is no longer available in MORNINGSTAR. Given the persis- tence of this practice, it is possible that fund managers have other reasons for closing funds that are unrelated to the performance of the fund. Performunce of Mutual Funds 269 NOTES 1. Tables detailing the results of the three-year comparisons are available from the authors upon request. REFERENCES Barber, J.R. (1994). Mutual fund risk measurement and future returns. Quarterly Journal ofBusiness and Economics, 33(Winter), 55-64. Brown, S.J., & Goetzmann, W.N. (1995). Performance persistence. Journal of Finance, 50,679-698. Carhart, M.M. (1997). On persistence in mutual fund performance. Journal of Finance, 53, 57-82. Damato, K., & Jereski, L. (1997). Fidelity’s 34-day offer to jump in to Magellan Fund probably should be ignored, many advisers assert. The Wall Street Journal, August 28, C 1. Elton, E.J., Gruber, M.J., & Blake, C.R. (1996). The persistence of risk-adjusted mutual fund perfor- mance. Journal of Business, 69, 133- 157. Fabozzi, F.J., Francis, J. C., & Lee, C.F. (1980). Generalized functional form for mutual fund returns. Journal of Financial and Quantitative Analysis, 15, 1107-l 120. Fabozzi, F.J., & Francis, J.C. (1979). Mutual fund systematic risk for bull and bear markets: An empirical examination. Journal of Finance, 34, 1234- 1250. Gibbons, J.D. ( 1985). Nonparumetric methods for quantitative unulysis. Columbus, OH: American Sciences Press, Inc. Goetzmann, W.N., & Ibbotson, R.G. (1994). Do winners repeat? Patterns in mutual fund return behavior. Journal of Portfolio Management, l&Winter), 9- 18. Grinblatt, M., & Titman, S. (1992). The persistence of mutual fund performance. Journal ofFinance, 47, 1977-1984. Gruber, M.J. (1996). Another puzzle: The growth in actively managed mutual funds. Journal of Finance, 51, 783-810. Hendricks, D., Patel, J., & Zeckhauser, R. (1993). Hot hands in mutual funds: Short-run persistence of relative performance, 1974-1988. Journal of Finance, 48, 93-130. Jensen, M.C. (1968). The performance of mutual funds in the period 1945-1964. Journal ofFinance, 23, 389-416. Kahn, R.N., & Rudd, A. (1995) Does historical performance predict future performance?. Financial Analysts Journal, Sl(November-December), 43-52. Lehmann, B.N., & Modest, D.M. (1987). Mutual fund performance evaluation: A comparison of benchmarks and benchmark comparisons. Journal of Finance, 42, 233-265. Malkiel, B.G. (1995). Returns from investing in equity mutual funds 1971 to 1991. Journal of Finance, 50, 549-572. Phelps, S., & Detzel, L. (1997). The nonpersistence of mutual fund performance. Quarterly Journal of Business and Economics, 36(Spring), 55-67. Sharpe, W.F. (1966). Mutual fund performance. Journal of Business, 39(Special Supplement), 119- 138. Treynor, J.L. (1965). How to rate management of investment funds. Harvard Business Review, 43(January_February), 63-75.