ETF Tracking Error: 'Need' vs. 'Demand' Funds

by: Herb Morgan

Herb Morgan (Efficient Market Advisors, LLC) submits: As the ETF industry expands its reach, marketing the product creations can stretch in conjunction. While this observation is not profound it is well understood within the fraternity of mutual fund industry professionals. Need and demand are two very different things when it comes to financial products in general and Exchange Traded Funds in particular.

I define Need as empirically provable within a range of statistical relevance. Investors need certain asset classes in their portfolio if they wish to optimize the relationship between risk and return. I call the need asset classes “core.” Core asset classes are Large Cap Stocks, Mid Cap Stocks, Small Cap Stocks, Large Cap International, Short Term Investment Grade Bonds, Intermediate to Long Term Investment Grade Bonds, Cash, Real Estate Investment Trusts, and Commodities.

If an investor adds to a portfolio beyond the core asset classes with demand assets, or what I call “non-core” assets, there is a certainty that risk waxes within a portfolio without a statistically provable corresponding increase in return. Non-core asset class Exchange Traded Funds include: Micro Cap ETFs, Emerging Markets ETFs, Single Country ETFs, or Narrow Sector ETFs.

Having spent the majority of my professional life as a senior officer of some of the nation’s largest mutual fund complexes and brokerage firms, I have seen first hand the tweaking of empirical research to create a need for what arguably should be a demand product. Mutual Funds and ETFs are sold rather than bought. Manufacturers know this, which is why they spend copious amounts of money on distribution. The idea here is to create a perceived need with their target audience, Retail Stockbrokers and Financial Planners. The target audience is a willing participant in the process in order to justify commissions and fees by complicating the investment process.

One would expect a certain amount of tracking error from ETFs due to fund expenses, however “demand” ETFs have experienced tracking error well in excess of a statistically tolerant range. The 2005 report card for need ETFs and demand ETFs is in, and predictably demand ETFs had significant tracking error. International ETFs had an average tracking error of 103 basis points, US Sector & Industry ETFs had tracking error of 62 basis points, and Global ETFs had checked in at 76 basis points. This compares with 2005 tracking error in need ETFs of 18 basis points for US Major Market ETFs, 18 bps for US Fixed Income ETFs and 23 bps for style ETFs.

Tracking Error in Exchange Traded Funds occurs for a number of reasons. One obvious culprit is costs, which can be expected. ETF manufacturers tend to charge more for demand products than need products which contributes to a portion of the tracking error. Beyond that there are diversification issues that stem from the inability of a fund to replicate an index and still maintain status as a mutual fund. Mutual funds may not have more than 25% of their assets in any given stock.

This presents a problem in the case of ETFs such as the IShares Dow Jones US Energy Sector (NYSEARCA:IYE) where Exxon Mobil (NYSE:XOM) makes up 29% of the index. The IShares Dow Jones Select Telecommunications (BATS:IYZ) and the Vanguard Telecommunications Service VIPERs (NYSEARCA:VOX-OLD) have three stocks which represent 68% and 65% of their target indexes respectively. Beyond the 25% single stock limit there is also a requirement that the sum of the weightings of 5% or greater positions cannot exceed 50% of the assets of the fund. In 2005 the IShares MSCI Brazil Fund (NYSEARCA:EWZ) had 2005 tracking error of 399 basis points because two stocks Petroleo Brasilerio and Vale do Rio Doce made up over 50% of the index and significantly outperformed the index.

Finally, there are liquidity issues which can contribute to tracking error. One example is the IShares Russell Micro Cap Index Fund (NYSEARCA:IWC). Clearly, there is a lack of liquidity in the micro cap space. It has been suggested that the Russell Company may have made changes to the index in order to accommodate the product folks at Barclays, who pay handsomely for the use of an index. There are companies in IWC with market caps in excess of $1 billion. I have always defined Micro Cap as companies with market caps below $500 million. The Russell Micro Cap Index consists of the bottom 1000 companies of the Russell 2000 Index and the next 1000 companies by market capitalization. Given the lack of liquidity in the bottom 1000 companies, Barclays is force to “optimize” the portfolio. Optimization is a code word for buying more liquid companies which in this case are a fairly sizable number of regional banks.

Manufacturers of demand ETFs really cannot improve on their tracking error. For investors, the issue to consider is whether such asset classes actually belong in a portfolio. Throughout my career, demand asset classes have blown up time and again usually at the height of the acceleration of the sales trend. Remember the Peso devaluation of 1994, or the Asian crisis of 1998, the tech blowup of 2000? There will be unforeseen circumstances over the coming years that will cause these products to fail unexpectedly and quickly. When this happens gains are quickly erased. Investors should stick to prudent asset allocation of core asset classes, shut down their charting programs and go fishing. Please visit our website to download a free copy of my book, “Good Money – Bad Investments”.