A reader sent in the following question in response to The Seeking Alpha ETF Investing Guide: "I was reading William Bernstein's argument that index mutual funds will outperform equivalent ETFs because of the lower tracking error, even though they do have higher expense ratio. What do you think of that article?" My answer:
William Bernstein is definitely worth listening to, and his book The Intelligent Asset Allocator worth reading. In fact, I include his web site as a valuable source in The ETF Resource Page. But I disagree with him about ETFs, and it may not surprise you to hear that the criticism he cites is actually from my review of his book on Amazon. (If you're interested in reading it, the review is titled Where Was the Final Chapter?, and actually cites my web site in its conclusion.)
Here's where I disagree with Dr. Bernstein. Most people who read the academic literature and the literature for investment management professionals know that active management trails indexing, and that portfolio composition (ie. asset allocation) determines the majority of investment returns. On that we all agree. So investors should focus on asset allocation, and implement that asset allocaton with index funds.
The specific index funds you choose should then be determined by two factors: cost and ease of portfolio management. The cost side includes spreads, commissions, tracking error, annual fees, and tax efficiency. I'd be perfectly willing to consider that Vanguard funds could/have/will beat ETFs on that basis. I don't work for Barclays or State Street, so I have no financial interest or emotional attachment that weds me in principle to ETFs versus index mutual funds.
However, the most important part of The Seeking Alpha ETF Investing Guide was not its presentation of the arguments for asset allocation and low cost indexing. Rather, it was the observation that "Buy and Hold" does not make sense when asset classes are swinging around as they have over the last 5 years (and for that matter much longer). Individual investors, I argued, should follow the same strategy as the most astute professional investment managers: they should rebalance their portfolios to reduce risk and boost returns. The impact of portfolio rebalancing - particularly if you can offset the resulting capital gains with losses for tax purposes - is considerable. (Here's a summary of the section on How to Manage Your ETF Portfolio that discusses rebalancing and tax-loss selling.)
So the key question for individual investors is: what instruments can I use that (a) minimize my costs and (b) maximize the ease with which I can rebalance my portfolio while minimizing taxes? And the answer, I think, is clear: for individual investors, ETFs win outright because you can easily find numerous alternatives for tax loss selling, you can use limit orders to help you rebalance, and ETFs are intrinsically more tax efficient.
If you use Vanguard funds, how do you take advantage of a decline in your holding of the small cap index? There's no alternative fund to switch to for tax loss realization. But if you use ETFs, you can switch from the small cap 600 to the Russell 2000 and back again, staying fully invested but realizing tax losses to offset your gains. The advantage is dramatic, because being able to realize tax losses potentially allows you to rebalance without paying a tax penalty. In contrast, using traditional no-load mutual funds tends to lock you in to a single provider (such as Vanguard) with no alternatives for tax loss selling.
In fact, Dr. Bernstein, whose first book was written before ETFs and who fails to discuss how to offset rebalancing with tax-loss selling in his second book, himself states that investors should forget rebalancing in taxable accounts because of the tax penalty. Well, if you failed to rebalance between 1996 and 2004 as the stock market rocketed and then swooned because you were using Vanguard funds, your losses compared to a rebalanced, tax-managed ETF portfolio would make the tracking errors Bernstein discusses look insignificant.
Why is it that Bernstein focuses only on costs and not on ease-of-management? My suspicion is that it's all down to personal experience. William Bernstein runs his own investment advisory firm, and has access to Vanguard Funds and competing index funds from Dimensional Fund Advisors. He can switch between the two for tax-loss selling. But Dimensional funds are not available to individual investors, so you can't.
If you want William Bernstein to replicate his experience (and access to Dimensional funds) with your money, you have to pay him. And most managers like him charge asset-based fees which eat a percentage of your total assets each year. If you want to know what I think of asset-based fees and what they'll do to your investment returns, read this.
Perhaps for the same reason (namely that he runs a professional investment advisory firm), Bernstein lacks appreciation for the convenience of ETFs for regular investors. Most investors hold some sort of stock or mutual fund account already. ETFs allow them to keep their accounts and move to an indexing strategy. Forcing people to move their accounts to Vanguard (remember - you can't buy Vanguard funds in many fund supermarkets or competing mutual fund accounts) is simply impractical. Instead of supporting practical ways for investors to move to indexing and asset allocation (which he has been a great advocate of), Bernstein has instead painted himself as a Vanguard fanatic. What a pity.
As for the particular tracking-error statistics Bernstein quotes, I think it's early days to crown a winner. ETFs are newer than index mutual funds, and their fund managers are improving with time and as assets in ETFs grow. (Yup, ETFs are gaining market share.) For example, many ETFs had capital gains distribution in their early years, whereas in 2003 Barclays had zero capital gains distributions in its ETFs.
More important, it seems that ETFs operate in a more competitive market than index mutual funds, and with time that will drive down their costs and increase their efficiency. Here's why the ETF market is more competitive: as an investor, you can choose between all the available ETFs from every ETF provider when you decide to purchase one. But with index mutual funds, you are sometimes locked into the products of a single provider if you have your account with that firm.
We're already seeing the benefits of extreme competition in the ETF market. Over the last twelve months, for example, SPY has lost market share to IVV, though (a) both are index ETFs that track the S&P 500, (b) SPY is a larger and more liquid fund, and (c) SPY has a longer operating history. The reason for SPY's loss of market share? Probably that SPY's 0.12% annual expenses were significantly higher than IVV's 0.09%. (Though note that both are lower than the Vanguard 500 Fund's annual expenses of 0.18%.) State Street responded by cutting the expense ratio on SPY by almost 17% to 0.10%. Over time, competition should continue to drive down ETF fees, and should also raise the bar on tracking error, tax efficiency, and spreads.
In fact, Vanguard now offers ETFs for its index funds. So if Barclays iShares are less efficient (including tracking error) they will loose assets to Vanguard ETFs over time. And as if to prove that greater competition exists in the ETF market, Vanguard's Total Stock Market VIPER (an ETF) has lower expenses than the identical Vanguard Total Stock Market Index Fund (0.15% versus 0.2%). VIPERS, by the way, are a share class of the existing Vanguard funds, so the tracking errors of the ETF and index fund are identical.
Greater competition in the ETF market should also lead to broader choice than is available with index mutual funds. ETFs for gold bullion (a glaring ommission from most portfolios) are already in registration; yet there is no open-ended pure bullion mutual fund available. I expect we'll see ETFs for other "missing" asset classes too, including foreign bonds and foreign real estate. In A Better Way to Invest, I argued that investors (including the wealthy) should adopt commodity products facing the stiffest competition, and that suggests ETFs rather than index mutual funds.
Finally, I think Dr. Bernstein underplays the greater tax efficiency of ETFs, which could be highly significant in future. Think of it this way: if you buy a Vanguard fund now and you're 30 years old, a lot of baby-boomers are going to cash in their shares before you do. When they do the market may be a lot higher than it is now. You risk getting a large capital gains distribution when that happens, crippling your compounded long run return.