ETFs Finish In First Place

angled in the morass of war in the 1960s, President Johnson proclaimed that one of his administration's goals was to "win the hearts and minds" of the Vietnamese people. To Mr. Johnson's chagrin, however, large blocs of the Vietnamese populace were singularly unswayed by American blandishments. It's tempting to say that the purveyors of exchange-traded funds set a similar course for themselves when they floated rafts of new funds over the past few months. Financial pundits readily lined up on either side of the launching ramp, too -- some to decry ETFs' variances from net asset values and their attendant transaction costs, others to laud the products' tax efficiency and low holding expenses.

Certainly, ETFs represent a growth industry. ETF net assets grew 20.2% last year (through November 30), according to the Investment Company Institute. Over that same period, the ICI reported, stock mutual fund assets suffered a 15.5% decline. Admittedly, the ETF asset base is much smaller. ETFs have been around less than a decade, and ETF assets amounted to less than 2% of stock mutual funds' at year-end 2000. Then again, mutual funds have had nearly a 70-year head start over ETFs in acquiring customers.

Anecdotal information abounds about financial professionals embracing ETFs, but the question remains: How have ETFs measured up? Have they delivered on their promises of asset-class purity, tax efficiency and lower costs?

Champions in the battle for the hearts and minds of planners are the fund houses claiming index primacy. In 1973, Barclays Global Investors (then the trust department of Wells Fargo Bank) brought out the first S&P 500 Index fund for institutional customers. Three years later, the Vanguard Group bestowed the same favor on retail customers with its introduction of the portfolio now known as the Vanguard 500 Index fund. Today, the firms square off, Vanguard with mutual funds and BGI with ETFs, over seven indexes: the S&P 500, together with its growth and value style sub-indexes; the S&P MidCap 400; the Russell 2000; and the S&P SmallCap 600 growth and value splits.

The year preceding the Sept. 11 terrorist attacks, while disappointing to many investors, can at least be said to be "normal" in terms of volatility, so it's worth a look at how the fight shaped up in the fund arena over the last normal year: August 2000 through August 2001.

Asset Class Purity

There are many ways to discern how well a fund cleaves to its benchmark. One method simply tracks the variations in returns between a fund and an index. Tracking error is then gauged as the standard deviation of these variances. Planners should be careful regarding tracking error, though. Measuring the differences from net asset value (NAV) is one thing; against total return (TR), it can be something altogether different. Take the variations against the S&P MidCap 400, for example. Vanguard's Mid-Cap Index Fund's NAV tracking error was 0.47%, but the deviation in daily returns shrank to 0.13% when dividends and capital gains distributions were reinvested.

In the aggregate, ETFs seem to track index returns significantly better than mutual funds on an NAV basis, but their advantage contracts considerably when total returns are used as the basis of comparison (Figure 1).

Not surprisingly, R-squared correlations follow a similar pattern. The R-squared portfolio statistic represents the degree of fund movement explained by changes in the benchmark index. The closer R-squared is to 1, the better the index describes fund behavior and the more trustworthy the fund's beta correlation.

Beta measures the volatility of a fund in relation to its benchmark. A beta correlation of 1 indicates that the fund is as volatile as the index; correlation figures above 1 signal greater volatility than exhibited by the benchmark, while numbers below 1 indicate relative quiet.

Performance

Some planners may find a little tracking error digestible if better-than-benchmark performance can be wrested from a portfolio. Just how do ETFs stack up against mutual funds in the performance category? A simple barometer compares fund absolute returns (NAV and total) against that of the benchmark index. A more sophisticated approach, alpha, measures the residual risk of a mutual fund in relation to the market. Think of positive alpha as the extra return awarded for investing in a specific fund instead of accepting the benchmark return. For example, an alpha of 0.02 means the fund outperformed the market-based return estimate (derived by using its beta) by 0.02%. Conversely, an alpha of -0.01 means a fund's return was 0.01% less than what would have been predicted from the change in the benchmark.

Even though three ETFs produced higher total returns, the performance edge, on average, was held by mutual funds (Figure 2). It's the wide disparity in the NAV and total return numbers generated by mutual funds, however, that's most telling. Mutual fund investors hoping to earn index returns seem obliged to reinvest their dividends and capital gains distributions, while those invested in ETFs can take close-to-index returns and still take cash distributions. On a volatility-adjusted basis, however, the contest is a draw, with each fund type averaging a 0.01 alpha.

Tax Efficiency

The impact of taxes on mutual fund returns was made exquisite for investors during the recent equity market slide. Many unfortunate mutual fund holders were handed tax bills while they watched their share values plummet. Capital gains distributions, wrought by other investors' fund redemptions, were often the culprits. Most mutual funds hold relatively small cash reserves, so a flood of shareholder redemption requests can force them to sell off portfolio securities in order to raise cash. Unless a fund is held in a tax-deferred account, any resultant tax liability gets divvied up among the remaining shareholders along with the capital gains realized from the sales.

ETF holders, in contrast, are insulated from the effects of other investors bailing because their transactions don't take place at the portfolio level. Instead, ETF shares are swapped between retail investors in exchange transactions. That's not to say that there aren't liquidations from ETF portfolios; there are. But ETF portfolio sales tend to be infrequent because they're usually limited to index rebalancings or portfolio optimizations. The creation and redemption of ETF shares through the portfolio is done only in large-block "in-kind" transactions by institutional accounts. And because these trades are a kind of barter, there are no capital gains consequences.

Holders of ETFs seem to suffer less, across the board, at the hands of the taxman compared with their mutual fund-holding brethren (Figure 3). Quantifying tax-efficiency, however, is possible only for the fund set that produced positive returns for the period; that is, those based upon the S&P SmallCap 600/Barra Value Index. Tax efficiency for the ETF version was 97.15%; the analogous mutual fund cranked at an 81.23% tax-efficiency rate.

Costs

Fund investments can nick investors in a couple of ways. First, there are entry, or transactional, costs. Mutual fund investors, for example, might encounter sales charges for broker-sold portfolios or fund supermarket transaction fees. Buying a fund directly from a distributor like Vanguard avoids such brokerage-related charges, but still may not offer the investor a free ride. For example, direct investors in Vanguard small-cap growth and value funds are charged a 0.50% purchase fee by the funds. ETF investors, on the other hand, are typically obliged to pay brokerage commissions unless their funds are held in "wrap" or asset management accounts.

Then there are the holding costs associated with fund shares. The most obvious costs are ongoing management expenses, but small shareholders in Vanguard funds may also be charged account-maintenance and low-balance fees as well. Wrap account fees can attend both ETF and mutual fund holdings.

In an apples-to-apples comparison, ETF operating expenses are lower, on average, than even bargain-basement mutual fund charges. But it's in the area of opportunity costs that debate is more likely to swirl. Like stocks, ETFs are traded in continuous auction markets subject to bid-ask spreads. Spreads, together with the possibility of variations from NAV, represent a price for the immediacy of an ETF transaction.

Specialists, as part of their commitment to the exchange, promise to keep their spreads to a certain width. However, a sizable number of ETF trades (47.5% as observed here) take place inside the spread. A simple function simulates an "effective spread" by multiplying the specialist's promised spread by the percentage of trades not executed inside. For example, if a spread is promised no wider than 0.10%, but 40% of trades are observed to take place inside, the effective spread becomes 0.06% [0.001 x (1 - 0.4)].

Variances from NAV exhibited by ETF share prices can represent either a windfall or an expense. ETFs displayed, on average, a modest end-of-day discount from NAV. For an ETF buyer, such a discount is clearly advantageous since a dollar's worth of assets can be effectively bought for only 96 cents. If, however, the discount widens by the time a sale is contemplated, a cost results. Netting the absolute value of the average premium/discount together with the effective spread yields a total opportunity cost for the ETF.

Mutual fund investors face opportunity costs, too. The delay in executing their fund orders until day's end represents a risk assumption since intra-day market volatility can either award a boon or impose a cost. To quantify this risk, an average investor's risk aversion factor, or lambda, posited as 10, is multiplied by the square of a fund's standard deviation of daily returns.

Before considering commissions, supermarket fees and other front-end costs attending portfolio investments, ETFs seem to hold an edge over mutual funds in the cost category (Figure 4).

Planners long accustomed to employing mutual funds in their portfolio allocations may look upon these findings with a shrug. For those managing money in tax-deferred accounts especially, ETF tax efficiency may resound only with a thud. Still, the empirical data indicate that ETFs, on balance, seem to have matched the claims of lower costs, better index tracking and greater tax efficiency made during the products' rollouts. Promoters of ETFs can only hope that by leading with the minds of planners, hearts will follow. — Brad Zigler

_________________________________________________________________________________
Brad Zigler is a founding member of the Global Association of Risk Professionals Education Committee and former manager of options marketing, research and education at the Pacific Exchange. He can be reached at brad_zigler@hotmail.com.