Why fees are more important than ever.
The case for low-cost investing is stronger in a world of low expected returns and increasing sophistication, and index funds and exchange-traded funds are a great way to keep expenses to a minimum.
The financial crisis and its echoes have led many investors to revise their investment strategies. The two major thrusts among institutional investors have been either to search harder for funds that may outperform or to hold more low-cost index funds. Most investors and advisors will be better off adopting the low-cost approach as the quest for a shrinking market-beating pie becomes more brutal and the penalty for mistakes becomes even steeper. ETFs and other passive vehicles are great ways for investors to earn a fair share of the slower-growing market pie.
In order to boost low consumer demand, central banks have driven down real interest rates to punishingly low levels. The theory is that cheaper capital will encourage businesses to invest and banks to make loans, and inflated asset valuations will make consumers feel richer and more open with their wallets. The banks have succeeded in that risk-asset prices have risen and expected returns have fallen.
Low expected returns effectively increase asset-based fees such as expense ratios. Another way to think about fees is as a percentage of an asset's or a strategy's expected return. Through this lens, a 1% annual charge is egregious for an asset expected to return 2% annually--half the expected returns are consumed--but reasonable for an investment expected to return 10%. However, asset-based fees are sticky, so throughout a market cycle, a strategy's costs as a portion of expected returns can vary dramatically. When markets are overheated and therefore likely to underperform going forward, investors paradoxically are willing to pay higher fees and to believe that fees as a portion of future returns are de minimis. The mechanism works in reverse when valuations are depressed.
Many investors are unaware that an asset class's long-term expected returns can be reasonably estimated by summing the asset's carry (or yield) and the carry's expected growth. (We are ignoring changing valuations, as they wash out over time and are driven by sentiment.) Over long periods, dividend growth per share has been stable for most economies at about 1% a year in real terms, allowing for surprisingly predictive long-run return estimates for equities. Similarly, yield-to-maturity and historical default rates do a decent job for bonds. For almost all asset classes, current yields suggest a low-return world going forward, perhaps one of the reasons why managers like PIMCO's Bill Gross and GMO's Jeremy Grantham are gloomy about long-run risk-asset returns.
As of writing, the Barclays Capital U.S. Aggregate Bond Index's yield to maturity is 2.5%, implying a subnegative 1% forward-looking real return. With such slender returns, fees become even more important. Vanguard Total Bond Market ETF
A Look Back at Large-Blend Equities
A number of ETFs have accumulated 10-year performance track records, a long enough time frame to lend some statistical confidence on the debate on index mutual funds versus ETFs. Still, there are challenges to conducting long-term studies of funds that necessitate the use of some assumptions. For example, there is a survivorship bias if we only take into account funds that survive the entire period because the weaker funds tend to disappear. But investors care about risk and returns, not an arbitrary success ratio. Was the upside in the 22% of the funds that beat the S&P 500 worth the risk of ending up in one of the 78% of the funds that did not?
To answer this question, we need some way to deal with the large number of funds that did not survive the study period. One way to deal with survivorship bias is to substitute a return for the funds in the months after they disappear. We could assume that investors in closing funds reinvest in the index or receive a return equal to the average of all of the remaining funds. Using the index return is perhaps the most generous to the disappearing funds because it does not include any expense ratio drag and it ensures the fund will at least match the benchmark after it disappears.