Comparative Performance of Load and No-Load Mutual Funds in Canada, The

Canadian Journal of Administrative Sciences, Dec 2004 by Deaves, Richard

Abstract

This paper investigates, using both single-factor and multi-factor models, the absolute performance of Canadian equity funds and the relative performance of load versus no-load funds. Consistent with a wealth of other studies, I find that the typical fund manager in Canada is unable to surpass his risk-adjusted benchmark. Moreover, the advantage possessed by load funds, in being able to undertake fewer liquidity-motivated trades than most no-load funds, does not translate into their being able to outperform no-load funds, even when loads are ignored.

Résumé

Le présent article utilise les modèles de facteur unique et les modèles de facteur multiple pour examiner la performance absolue des fonds d'actions canadiens et la performance relative des fonds avec frais d'acquisition, par opposition aux fonds exempts des frais d'acquisition. Comme les nombreuses études antérieures, notre recherche débouche sur la conclusion qu 'au Canada, le gestionnaire de fonds type est incapable de surpasser son point de référence ajusté en fonction du risque. Par ailleurs, l'avantage lié aux fonds avec frais d'acquisition, notamment sa capacité à entreprendre moins de transactions nécessitant des liquidités que les fonds exempts des frais d'acquisition, ne se traduit pas en capacité à donner de meilleurs résultats que les fonds sans frais, même si on ne tient pas compte des frais.

The potential ability of money managers to add value is one of the most heavily researched issues in finance. The early seminal paper was by Jensen (1968), who concluded that managers were unable to earn riskadjusted returns, even before consideration of expenses. Some later results were less disheartening for the pros. Ippolito (1989) and Grinblatt and Titman (1989) found that managers were, in fact, making a contribution in that they were able to recoup for investors the expenses charged. Unfortunately, it is now evident that survivorship bias taints many of these early studies. After correcting for this bias, Malkiel (1995), Elton, Gruber, and Blake (1996a), and Gruber (1996) once again concluded that, on average, managers were not able to surpass passive risk-equivalent benchmarks net of (and sometimes even gross of) all expenses.

Research along these lines in the Canadian context has also been conducted, most notably by Berkowitz and Kotowitz (1993); Kryzanowski, Lalancette, and To (1994, 1997); Athanassakos, Carayannopoulos, and Racine (1999); and Deaves (2004). Of these, only Berkowitz and Kotowitz (using a 1977-86 sample not free of survivorship bias) concluded that Canadian equity fund managers were able to earn excess returns after accounting for all expenses and loads. To focus on Deaves (2004), which has the advantage of utilizing a sample that is carefully constructed so as to avoid all bias, in line with the preponderance of U.S. research, this paper too finds significantly negative performance. Still the negative performance does seem to go hand in hand with some valued added from portfolio management, suggesting that some though not all of expenses charged are recouped.

Categorical conclusions on performance remain elusive. There are three reasons that I wish to focus on. First, in order to calculate risk-adjusted returns it is necessary to possess the right model. All students of finance understand we are not there yet. While the earliest studies utilized a single-factor (e.g., capital asset pricing model - CAPM) approach, it is now commonplace to condition on multiple sources of risk. This does matter, since it is possible to conclude superior performance using one benchmark but inferior performance using another. For example, Elton, Gruber, Das, and Hlavka (1993) found that the favourable results of Ippolito (1989) evaporated once a benchmark was utilized that was more representative of the actual universe available to fund managers.

second, is it possible that, while the average manager is unable to add value, there are a few strong performers who consistently do so. Studies by Grinblatt and Titman (1992); Hendricks, Patel, and Zeckhauser (1993); Goetzmann and Ibbotson (1994); Gruber (1996); and Elton, Gruber, and Blake (1996b) indeed found convincing evidence that successful performance was not purely random. Managers who were able, in a given period, to outperform risk-adjusted benchmarks were more likely than not to repeat their success. While this evidence has been questioned on a number of technical grounds (see, for example, Brown, Goetzmann, Ibbotson, & Ross, 1992; Carhart, 1997), some have remained convinced. Gruber (1996) argues that the choice made by some investors to entrust their money to successful active managers (Sirri & Tufano, 1998) may be a judicious decision given that such "new money" has in the past been able to outperform the market. In Canada a similar conclusion seems merited. Deaves (2004) presents evidence that, at least on a short-term basis, success breeds success, and investors seem aware of this since money flows to successful funds.