ETF Tax Efficiency Part 2 - In Defense of Index Funds By John Spence May 9, 2002

Yesterday we posted an article on exchange-traded fund (ETF) tax efficiency. We looked at how ETFs function and why proponents claim they will be more tax efficient than traditional index mutual funds over time. We also posted data from ETF researcher Brad Zigler: capital gains distributions and after-tax performance of Vanguard index funds and Barclays Global Investors iShares that track seven indexes.

Morningstar senior fund analyst Scott Cooley, who covers Vanguard funds, explained why the time period examined (August 2000-August 2001) may have skewed the results in favor of the ETFs.

IF: Have redemptions forced Vanguard index funds to make capital gains distributions?

SC: OK, the claim is that redemptions from open-end index funds have caused the capital gains distributions. In fact, at several of the funds that have gone into net redemption, Vanguard has actually booked tax-loss carry forwards by selling relatively high-cost lots of shares. In this regard, the traditional fund format can be a bit of an advantage because those net loss carry forwards can be used to offset future gains that are incurred because of rebalancing.

IF: Why did the Vanguard funds make distributions over this particular time period?

SC: The real, primary reason some of the Vanguard funds have made capital gains distributions is stocks that have often been long held were sold out of the funds in 2000 because they graduated into a larger-cap index or moved from a value index into growth. This increasingly becomes a problem for value- and small-cap-oriented indices because in an environment in which stocks post gains over time, the number of low-cost-basis shares they hold in the portfolios tends to rise. Most of the Vanguard style-specific index funds were also highly tax-efficient early on, but became less so as the bull market progressed in the 1990s. I do think the ETF format will allow Barclays to moderate these distributions over time, but I think the numbers used vastly overstate the ETF advantage in this case.

IF: How could fair comparisons be made?

SC: I think a more interesting, better-designed, and fairer study would have been to compare the distributions made by the two types of funds in calendar-year 2001, which was the first full calendar year of operation for many of the iShares included in the study. That also has the methodological advantage of showing how the funds would have performed in similar, down market environments. By using the time frame he employed, Zigler caught fiscal years for the Vanguard funds that include the tail end of the bull market, while the iShares were launched after the market peak.