Blowing smoke on trailer fees - MoneySense

Blowing smoke on trailer fees

Fees harm investors. Here are the facts


In the buildup to the regulatory review of trailer fees in Canada, the mutual fund industry is trying to wage an ill-advised battle of misinformation. And one of the key tactics in this battle is to dispute facts and studies, including one I co-authored, that have proven beyond a doubt the detrimental effect of these fees.

At the risk of making an analogy to the cigarette industry and early denial of the harm caused by cigarettes, I hope we stop blowing smoke and make use of the information and data provided by the mutual fund industry that clearly show trailer fees harm Canadian investors.

For investors who have not been following these developments, the Canadian Securities Administrators are calling for suggestions on how to value the price of advice when trailers, an embedded fee that compensates advisors directly for sales, are done away with. In a recent letter to the Ontario Securities Association, the Investment Funds Industry of Canada (IFIC) revisited its argument that studies have failed to prove harm.

Let’s take a closer look at the contents of that letter, sent regarding CSA Consultation Paper 33-404: Proposals to Enhance the Obligations of Advisers, Dealers, and Representatives toward Their Clients. On page 25, the IFIC letter states: “Limitations of Conflict of Interest Research: While it is important to consider the evidence cited in the Consultation Paper, it is also important to consider the limitations of the quoted studies.

“With respect to conflicts of interest research, a great deal of weight is put on aggregated fund-level analyses. This sort of research, using regression analyses, can identify interesting trends that may warrant further examination. For example, the Dissection of mutual funds fees, flows and performance (Cumming et. al., 2015) study demonstrates that compensation models may impact the flow-performance sensitivity.

“While this might be indicative of the presence of conflict it certainly does not prove conflict. On the fund level, the report found that the relationship – of better past performing funds generating higher sales – is less strong for funds that pay trailers compared to fee-based funds.

“However, because this analysis lacks portfolio-level data it is impossible to know if this relationship is harming or even benefiting investors in the long-run. A specific fund may play many roles in the overall client portfolio. Long-term longitudinal research would need to consider the role that a specific fund (and the associated compensation model) plays in the total client portfolio performance, risk management and compensation to conclude that a conflict exists.”

In short, IFIC is critiquing our paper “Dissection of mutual funds fees, flows and performance” (co-authored with Sofia Johan and Yelin Zhang at York University’s Schulich School of Business) by saying it is inconclusive due to the type of data we examine. While we strongly believe that IFIC has every right to question studies that may affect it as an organization and its members, we would be remiss if we didn’t further explain why this most recent IFIC critique is grossly mistaken and misleading.

Previously, IFIC engaged in a critique of our work by releasing a competing report that they commissioned from Investor Economics. Importantly, note that the data in the IFIC commission Investor Economics report uses the exact same type of data that IFIC criticizes in the Cumming et al., 2015 report. So, IFIC critiques the type of data used in our study while at the same time uses the same type of data to support their lobbying efforts when they think that conclusions warrant.

We previously commented on the empirical methods in the Investor Economics report in our comment letter on the OSC webpage entitled “Frequently Asked Questions about the Dissection of Mutual Fund Fees, Flows and Performance Report”, and as such I will not repeat those comments here. Instead, here I focus on the latest IFIC critique quoted above.

In this latest critique, IFIC is correct to point out that in our paper we do not look at specific accounts. We look at all of the flows into all of the different series/purchase option combinations provided by the industry to us. These flows into the series/purchase option combinations are aggregations of all investors’ accounts. As much as my coauthors and I would jump at the chance to analyze specific accounts of all mutual fund investors, our access to data is restricted by both practical and legal constraints.

If IFIC is able to convince every mutual fund investor to disclose the composition of their investment, retirement and education savings account, etc., we are happy to oblige by using the appropriate econometric methods for a new report. But it seems unlikely that any Canadian investor would provide permission for a fund manager to forward account level information to anyone for an empirical study due to the obvious confidentiality violations. IFIC’s suggestion to use investor account level data is therefore akin to suggesting the need for a study that is impossible carry out.

If we step back a minute, let me explain that the transfer of series/purchase option combinations level data for our study required severe confidentiality restrictions. Put differently, while the industry demanded that the regulators be evidence based, the industry was extremely reluctant to provide regulators with the means to do so. This reluctance was overcome by the industry’s recognition of the CSA’s determination to make a policy decision with or without the empirical research.

It was clear to that any decision about the future of trailer fees could potentially be made without first having an empirical study on topic (as per the approach in Australia, continental Europe, and the UK where trailer fees were banned). Given this context, it is rather surprising that IFIC is now suggesting the need to examine account level data.

Hypothetically, what would account level data enable? It would enable a precise examination of how much specific investors were harmed by trailer fees, and who exactly was harmed the most. But such account level data would not change the evidence from the series/purchase option combinations level data provided by the industry that clearly show that in aggregate investors are harmed by trailer fees.

The series/purchase option combinations level data provided by industry clearly show that the harm from trailer fees comes in a few forms:

1. Investors’ capital is steered towards funds that have higher trailer fees;

2. Investors’ capital is less likely to be taken out of funds with poorer performance among funds that pay higher trailer fees, and;

3. The associated reduced incentives for fund managers of funds that charge trailers to generate performance and capital flows is associated with lower future performance. For example, funds that raised their trailer fees experienced a drop in performance, while funds that lowered their trailer fees experienced a rise in performance.

My co-authors and I are acutely aware that the findings from our study are unpopular with the mutual fund industry (to put it mildly). We just keep reminding folks to please keep in mind that we have not offered any opinions or conjectural statements, and merely report what the data indicate, and have done so following standard statistical and econometric methods. The paper has been invited for presentation at a number of events both locally and abroad, including Europe and Asia, and was recently presented at the Northern Finance Association Annual Conference in September 2016 at Mont Tremblant.

We appreciate that the industry has a substantial financial interest in keeping trailer fees in Canada, with over $5 billion per year charged to Canadian investors. My co-authors and I have no financial stake one way or the other. We simply report what the data indicate. Blame the data. Please don’t shoot the messenger.

Douglas Cumming, J.D., Ph.D., CFA, is Professor and Ontario Research Chair, York University Schulich School of Business