The impact of fees on portfolios

By Esko Mickels | June 1, 2009 | Last updated on June 1, 2009
4 min read

If you held the fixed-income fund that had the worst performance in Canada over the last year, you’d likely be disappointed by its 3.4% loss. But you’d certainly be more upset if you held Canada’s worst performing high-yield bond fund (which lost 34.7%) or Canadian equity fund (down 45.4%). With these returns in mind, it seems clear a one or two percentage point difference in fees will impact a fixed-income fund’s return, but less so with equity funds.

Funds with higher fees tend to exhibit worse performance than those with lower fees. And because bonds typically generate lower returns than equities, their fees gobble up a greater proportion of their returns.

Imagine you have two investments. Before fees, investment A is expected to return 8% and investment B, 4%. If both investments cost 2%, A gives up a quarter of its expected return, while B sheds one-half of its expected return. Granted, fixed-income funds usually charge lower fees than equity funds, but the differential isn’t enough to make this point moot.

Even within the fixed-income space, fees won’t have the same effect on all fund categories. For instance, government bonds are expected to generate lower returns than high-yield bonds over time due to their relative safety. There is also less volatility and dispersion of returns in conservative bond categories. Hence, fees will have a greater relative impact on more conservative funds (see chart).

Typically, bond managers will bet on yield curve positioning, duration and/or credit quality. However, when managing a portfolio of Canadian government bonds, credit quality isn’t a factor. Instead, managers focus on varying the coupon and maturity of their bonds. But unless the manager is making big bets, or is continuously correct or incorrect, these tweaks tend to have a muted impact.

In examining the relationship between fees and returns, we compared a fund’s management-expense ratio (MER) to long-term returns (10 years) within various Canadian fund categories. We found the negative correlation between fees and returns was very strong for money market funds, strong for Canadian bond funds, and low—but still important— for equity funds.

These statistical relationships don’t prove causation, but when viewed with the economic rationale for low fees, they lend support to our belief that fees matter more for fixed-income funds than they do for equity funds.

In the United States, Morningstar’s Russel Kinnel observes in his new book, Fund Spy, the chance of a low-cost fund outperforming its category average is higher for bond funds than equity funds. U.S. stock funds in the cheapest quintile were 2.5 times more likely to beat their category average than funds in the most expensive quintile, while taxable bond funds in the cheapest quintile had a 6.9 times advantage.

There’s also a strong relationship between costs and survivorship, which compounds the negative attributes of high-cost funds. For instance, with U.S. taxable bond funds, Kinnel found funds in the most expensive quintile had a much greater rate of closures and mergers (57%) than those in the least expensive quintile (35%) over 10-year rolling periods since 1995. Not only are investors paying a lot, they also risk having their fund close on them.

The implication is when investors are evaluating funds, they should place greater weight on the fees they pay for fixed income funds than for equity funds. This means when selecting fixed income funds—especially moneymarket funds—investors may want to use low fees as one of their first screens. With equity funds, management, strategy, and stewardship gain importance in the selection process.

This doesn’t mean picking a low-fee fund will guarantee aboveaverage returns; far from it. But low fees do increase a fund’s chance of outperforming. A dollar paid in fees is a guaranteed loss.

The belief that higher cost equals higher quality is a myth. While it’s true some quality managers charge a high fee, it’s far from universal; great managers often charge low fees, and poor managers often charge high fees. There are examples of high-fee funds outperforming and low-fee funds underperforming, but the opposite is usually true.

The tough part is identifying managers whose skill and mandate combine to justify a higher fee. Over the last five years, only two out of 90 funds in the Canadian Money Market category surpassed the CIBC World Markets 91 Day Treasury Bill Index; no fund in the Canadian Fixed Income category outperformed the benchmark TSX DEX Universe Bond Index; and 19 out of 100 funds in the Canadian Equity category generated excess returns over the S&P/ TSX Composite Index.

The vast majority of funds fail to hurdle their category benchmark after fees, thus investors’ chances of blindly picking a benchmark beater are low. Consequently, investors need a high level of confidence that a fund will outperform to justify paying more. That or, they should be receiving additional benefits such as advice on risk management, tax and estate planning, budget management, or better communications from the fund company. These benefits are hard to put a price on, but thought should be given to their value.

Investors should also consider other important costs beyond the MER, such as fund trading costs, taxes and brokerage costs (for exchange- traded funds).

Passive funds generally have low fees, so they warrant attention, particularly in the fixed-income space. Investors seeking a passive, low- MER fund should consider the bond index ETFs in the iShares CDN DEX family (which have MERs between 0.25% and 0.4%) and Claymore family (0.15% to 0.25% MER), as well as TD Canadian Bond Index (e-series 0.48% MER), and RBC Canadian Bond Index (0.63% MER).

Alternatively, low-fee actively managed bond funds include: PH&N Total Return Bond (d-series 0.58% MER), PH&N Bond (d-series 0.59% MER), McLean Budden Fixed Income (0.65% MER), and Beutel Goodman Income (0.76% MER).

Esko Mickels is a fund analyst with Morningstar Canada.

Esko Mickels