Modestly improved investor education and extensive marketing have made investors today more conscious of the costs of their investment decisions, and helped promote the use of index-tracking mutual funds and exchange-traded funds as opportunities to reduce expenses and allow investors to keep more of their investment returns.
While cost-consciousness is a clear positive for investors, the move to indexing has created another source of risk investors rarely hear about: index construction.
How much exposure do you want to South Korea?
My favorite example of the risks of index construction is the FTSE and MSCI definitions of “developed” and “developing” countries outside the United States. They’re almost identical, except FTSE places South Korean equities in the “developed” bucket and MSCI includes them in the “developing” index.
This is fine if you’re invested entirely in funds from issuers tracking the same indices: iShares’ core developed (IDEV) and developing (IEMG) market funds both track MSCI indices, so South Korea’s inclusion in one and exclusion from the other doesn’t affect your exposure. Likewise, Vanguard’s developed (VEA) and developing (VWO) market funds both track FTSE indices, so if you only own Vanguard funds you’re still able to get exactly the exposure you want.
The flip side of that is someone who for whatever reason invests in a developing market fund provided by one company and a developed market fund provided by another. Owning a FTSE developed market fund and a MSCI developing market fund gives you exposure to South Korean equities in both funds, while owning a MSCI developed market fund and FTSE developing market fund leaves you without any exposure to South Korea in either fund.
I’m not trying to exaggerate the importance of South Korea to global capitalism (although South Korea has certainly played an important role in the development of global capitalism). I’m trying to say that if your goal is to invest in indexed mutual funds in order to reduce your need for ongoing supervision and maintenance, you need to put in more time up front to figure out exactly which indices your funds are tracking.
To put a little more meat on this bone, the MSCI-tracking IEMG has returned an average of 3.86% annually since October 2012, while the FTSE-tracking VWO has returned an average of 3.27%, while the MSCI-tracking IDEV has too short a history to be very interesting, but has slightly underperformed the FTSE-tracking VEA since inception.
Fidelity launched US and international index funds that track…something
That brings me to the news hook for this post: Fidelity made a splash this summer by introducing two funds with 0.00% expense ratios. The funds are only available to Fidelity customers, so you can’t buy them in Vanguard, Merrill Edge, or other brokerage accounts, but if you have some reason to invest through a Fidelity brokerage account, they’re certainly the cheapest funds available to you, charging as they do no fees on the amount invested in the funds.
Some people have suggested that these 0.00% expense ratio funds are being offered as a “loss leader,” but that’s not exactly right. In fact, they’re an experiment in charging people nothing in order to invest in Fidelity’s own bespoke indices, which they have the luxury of paying little or nothing for.
Here’s what Fidelity has to say about the “Fidelity ZERO International Index Fund:”
“Geode normally invests at least 80% of the fund’s assets in securities included in the Fidelity Global ex U.S. Index and in depository receipts representing securities included in the index. The Fidelity Global ex U.S. Index is a float-adjusted market capitalization-weighted index designed to reflect the performance of non-U.S. large-and mid-cap stocks.
“The fund may not always hold all of the same securities as the Fidelity Global ex U.S. Index. Geode may use statistical sampling techniques to attempt to replicate the returns of the Fidelity Global ex U.S. Index. Statistical sampling techniques attempt to match the investment characteristics of the index and the fund by taking into account such factors as capitalization, industry exposures, dividend yield, P/E ratio, P/B ratio, earnings growth, country weightings, and the effect of foreign taxes.”
And here’s what they have to say about the “Fidelity ZERO Total Market Index Fund:”
“Normally investing at least 80% of its assets in common stocks included in the Fidelity U.S. Total Investable Market Index, which is a float-adjusted market capitalization-weighted index designed to reflect the performance of the U.S. equity market, including large-, mid- and small-capitalization stocks.
“Using statistical sampling techniques based on such factors as capitalization, industry exposures, dividend yield, price/earnings (P/E) ratio, price/book (P/B) ratio, and earnings growth to attempt to replicate the returns of the Fidelity U.S. Total Investable Market Index using a smaller number of securities.”
Emphasis mine.
A bad fund can track a good index
Today, you could buy any of the following three ETF’s, all based on versions of the S&P 500 index:
- SPY tracks a market-cap-weighted S&P 500 index;
- RSP tracks an equal-weighted S&P 500 index;
- RVRS tracks an inverse-market-cap-weighted S&P 500 index.
All three options require you to decide whether the S&P 500 is a “good” index or a “bad” index. If it’s a bad index, you shouldn’t buy any of them!
But even once you’ve decided it’s a good index, you aren’t relieved of your responsibility. The existence of the three funds means you also have to decide what the best strategy is for investing in the companies that make up that index.
SPY has an expense ratio of 0.09%, RSP has an expense ratio of 0.20%, and RVRS has an expense ratio of 0.29%, so even if the funds have identical performance of their underlying assets, you’d be 0.11-0.2 percentage points per year ahead using SPY. However, in order to maintain their equal and inverse-market-cap weights, the latter two funds also have to do much more trading, incurring additional costs that aren’t reported in the expense ratio. That means your confidence in the investment strategy needs to be “somewhat higher” than the difference in expense ratios.
Conclusion
I’m sure Fidelity’s bespoke indices are fine, I’m sure their ZERO funds will track those indices very closely, and I’m sure they’ll perform similarly to the low-cost market-capitalization-weighted index funds offered by their competitors.
But my point is broader: you aren’t a passive investor just because you invest in low-cost, passively-managed mutual funds or ETF’s. You’re still making at least three very important, very active decisions:
- which indices to track,
- how to track them (market-capitalization, equal-weighted, inverse, leveraged, etc.),
- and how to allocate your assets between them.
You can carefully research those decisions, you can shrug those decisions off, or you can let me or the Bogleheads forums make them for you, but careful research, indifference, and reliance on the advice of others are all decisions too.
As the saying goes, it’s turtles all the way down.
jo says
heads up that scrolling on the home page isn’t working, i cant scroll down to the previous posts
calwatch says
The trading costs within the funds should be buried in the expense ratio. There are additional costs in that a fund like RSP or RVRS will probably distribute more capital gains simply because it has to rebalance more frequently, but the actual costs of trading are accounted for. The other issue is that since there are so many S&P 500 market cap weighted funds competition drives costs lower, while there is only one RSP and RVRS. I think you had mentioned on Twitter that the performance of RSP and a midcap ETF wasn’t much different but the midcap index fund was cheaper, so I sold my RSP and bought the midcap ETF instead. Even if the performance wasn’t that good, at least less of it went for profit for Guggenheim.
indyfinance says
calwatch,
Sorry I was writing quickly and wasn’t as precise as I try to be. It’s true things like broker fees and commissions should be in the expense ratio since they’re paid out in cash, but by “trading costs” I mean the costs of doing any trading at all, i.e. crossing bid-ask spreads in order to track a non-market-cap index. One of the key cost advantages of a market-cap-weighted fund is that if the index is broad enough the only trading that’s required is when issuing or redeeming shares. To track an equal-weighted or reverse-cap-weighted index, as RSP and RVRS try to do, they have to constantly buy underperformers and sell outperformers, which creates costs that aren’t included in the expense ratio.
As for RSP, I find the idea of equal-weighted funds intriguing, but not intriguing enough to put any of my own capital at risk in them.
—Indy