![Sample our Economics, Finance,Business & Industry journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5931/TOC-Subject-Sample-2021-Economics-Finance-Business-Industry.jpg"})
ABSTRACT
We investigate the smart money effect in the German mutual fund market from 2001 to 2016. Results show a positive relation between fund flows and subsequent performance for mutual funds with a European or international diversified investment focus. Funds that invest domestically, however, show no signs of a smart money effect. Moreover, evidence suggests that flows to funds managed by bank-affiliated investment companies are smart. We argue that less sophisticated investors rather invest domestically and that financial advice improves retail investors’ mutual fund investment decisions.
Acknowledgements
We are grateful to the editor, Chris Adcock, Christina E. Bannier, Andreas Walter, Hans-Hermann Francke, Thomas Heyden, and two anonymous referees whose constructive comments helped to improve this paper. We also thank the participants of the European Financial Management Association 2019 Annual Meeting in Ponta Delgada, Portugal, the VHB Conference 2019 in Rostock, Germany, and the Doctoral Consortium of the European Retail Investment Conference 2019 in Stuttgart, Germany, for helpful comments.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
1 According to the BVI dataset (see Section 4).
2 According to the BVI dataset (see Section 4).
3 The BVI is the German equivalent of the Investment Company Institute (ICI) in the US.
4 The average number of existing funds per year is 90 for funds with domestic investment focus, 452 for funds with European and 958 for funds with international investment focus.
5 In order to adjust for outliers, we truncate net cash flows (Flow) at the 0.5 and 99.5 percentile.
6 As mutual fund investor behavior could be different during times of crises, we re-estimate our multiple regression results (Table ) for a subsample of crisis periods. As we find no significant smart money effect during these periods, we conclude that our results are not driven by the two crises.
7 Following Zheng (Citation1999) we build above and below median flow portfolios in our main results as this approach accounts for overall trends in flows per quarter. In Table in the Appendix, we additionally report results for positive and negative flow portfolios.
8 We also use an equally-weighting scheme for new money portfolios (Table ). Unless otherwise indicated, we rebalance new money portfolio every quarter (3-month holding periods).
9 We estimate the variance inflation factors (VIFs) to test for potential multicollinearity issues resulting from using both control variables. The maximum VIF is 1.85, indicating no multicollinearity issues. Results are still consistent when controlling for only one of the lagged performance variables.
10 Please find results for equally-weighted portfolios as well as results for the positive and negative flow portfolios in the Appendix in Table and Table , respectively.
11 Using excess returns instead of alphas, the results on the smart money effect for funds with European and international investment focus remain qualitatively unchanged.
12 As there is evidence for the smart money effect in the US fund market (Keswani and Stolin Citation2008; Zheng Citation1999), we investigate whether the results for funds with international investment focus may be driven by investments in US equity. Using data on the asset allocation in US equity for each fund, we split this sample by the median in order to receive a sample with low and high US equity. The smart money effect is highly statistically significant for both subsamples and of comparable magnitude. Thus, the finding is not driven by investments in US equity. Results are unreported but available on request.
13 For example, Stotz (Citation2007) reports that actively managed mutual funds in Germany underperform their risk-adjusted benchmark by 1.9 percentage points p.a., whereas passive index funds underperform by 1.0 percentage point p.a. from 1998 to 2005. In this context, Blitz, Huij, and Swinkels (Citation2012) report that European index funds and ETFs underperform their benchmarks by 0.5 to 1.5 percentage points p.a. not only due to expenses but also due to dividend taxation.
14 The marginal increase in R2 through the Carhart four-factor model is reported by several studies (see e.g. Gharghori, Mudumba, and Veeraraghavan Citation2007; Vidal-García Citation2013).
15 We also investigate longer time horizons of the flow-performance relationship in a multiple regression framework. Results remain consistent when including fund flows lagged by two and three quarters.
16 Results of Tables , and were also estimated using the three-factor model for performance measurement. All results remain consistent and are available on request.
17 Several studies investigate the smart money effect separately for institutional and retail investors to analyze differences due to the higher sophistication level of institutional investors (e.g. Feng, Zhou, and Chan Citation2014; Jiang and Yuksel Citation2017; Keswani and Stolin Citation2008). In Table (Appendix) we compare our results for retail funds to a matching sample of institutional funds. For institutional fund flows, we find no dumb money effect for funds with domestic investment focus. For funds with European and international investment focus, we find a significant relation between fund flows and subsequent performance as well. This finding is in line with Keswani and Stolin (Citation2008), who report the smart money effect for both, retail and institutional funds.
18 The total shareholder costs are calculated as total expense ratio plus one-fifth of the front load.