Waddell & Reed

Market Perspectives

What you don’t know can hurt you: Understanding the potentially concealed cost of ETFs

Story Highlights

  • According to a New York University study, ETFs often trade at prices that are a premium or discount to their Net Asset Value (NAV).
  • Bid-ask spreads may vary widely based on numerous factors.
  • As with any investment, it is important to understand the potential risks.
Download PDF

The U.S. exchange-traded fund (ETF) industry marks its 20th anniversary this year, dating back to the Jan. 29, 1993, launch of the Standard & Poor’s 500 ETF (under the ticker SPDR) on the American Stock Exchange. Today, there are more than 1,200 U.S.-domiciled ETFs1 — up from about 100 a decade ago.

While a key marketing point for ETFs is that they are cheaper than traditional actively-managed mutual funds, the low price has the potential to ultimately carry a steeper cost than some ETF investors may realize. And those costs can be influenced by a range of factors, including some that investors might see as only tangentially related to their actual investment.

It is important to recognize that ETFs, by their nature, are not necessarily more risk-prone or more costly than other types of investments. All investing requires a measure of risk versus potential returns. Achieving investment goals often requires a wide ranging and diverse investment strategy to which ETFs can contribute.

ETF price = or ≠ NAV

In simplest terms, the goal of an ETF is to provide diversified and liquid exposure to a particular market index at a lower cost than traditional mutual funds. However, while traditional mutual funds price only once daily, ETFs trade throughout the day, creating a potential pricing risk related to how efficiently the ETF can respond to the movements of its individual components.

Most ETF investors would likely believe that the share price perfectly aligns with its net asset value (NAV), or the combined value of its holdings. That correlation is designed to occur through a self-correcting arbitrage mechanism where authorized participants (AP) — generally large institutional investors — can create or redeem ETF shares for holdings in the underlying portfolio. In large and highly liquid markets, such as the Standard & Poor’s 500 Index (S&P 500), that mechanism should — and generally does — function efficiently. The difference between an S&P 500 ETF and the fair value of its underlying stocks is often a penny or less.

But that spread can be significantly wider depending on numerous factors, including the market sector, investor activity and overall market volatility. A New York University study found that ETFs often trade at prices that are a premium or discount to their NAV2 which, of course, can act as a hidden cost on returns. For example, an investor buying an ETF with a 0.5 expense ratio — which is actually lower than the expense ratio on many ETFs3 — at a 1% premium would have an effective first year expense ratio of 1.5% — higher than many traditional mutual funds.

In many ways, the premiums relate directly to the holdings in the ETF’s underlying portfolio. As might be expected, the study found that the more substantial variances are more likely to occur in less liquid markets. For example, international equity ETFs, which can be hampered by time zone differences in the trading day, generally exhibited volatility of 50 to 130 basis points around NAV, while long-term municipal bonds and corporate bond ETFs, which involve over-the-counter trading and higher transaction costs, generally had volatility of around 50 to 160 basis points.

However, perhaps surprisingly, the study also found what might be considered significant variances in some more liquid markets as well. For example, some sector funds predominately based on U.S. equities had volatilities of 30 to 70 basis points around NAV. (It is also worth noting that similar ETFs do not necessarily move in the same direction relative to their NAV on the same day. The study noted examples of similar ETFs having end-ofday premium differences of +7% to -11% .).

It is also important to realize that some sectors may have a bias toward either a premium or a discount. For example, bonds are valued at their bid price instead of the higher ask because of fund accounting. The result: a NAV that is broadly viewed as understated, which means bond ETFs might be seen as trading at a premium. However, determining the actual value of bonds, which don’t have near the transparency of equities, can be especially difficult for an ETF investor. The difficulty can be compounded by the fact that high-yield bond ETF premiums can move substantially higher in periods of strong investor demand. Similarly, they can quickly turn into discounts when that demand falls. For example, investors fleeing high-yield bond ETFs during a rush out of that sector in November 2012 were selling ETFs at as little as 97 cents on the dollar. The run for the exits came less than a year after investors flocked to the space and pushed some junk bond ETF premiums more than 2.60% above fair value. As a result, investors who moved in and out of the related ETFs when flows were heaviest would have earned returns 3% below buying and selling the same securities on the same days at their fair value.4

ETF bid-ask and the space between

As is the case with any security, there is a bid-ask spread on ETF shares. For large and heavily-traded ETFs, the spreads may be as small as a penny per share. Or the spreads can be much wider — for example, several dollars per share — for ETFs that have fewer assets, that are thinly traded or that trade in less liquid markets. Depending on the market environment and the ETF, there have been cases where the bid has been less than one-half the ask.5 In that case, the ETF is a lot like buying a new car in that it loses a chunk of its value as soon as it trades hands. It is also important to recognize that the size of the spread does not necessarily correlate with price premiums or discounts. For example, an ETF trading at a high premium can still have an extremely small spread and vice versa.

Spreads can also be influenced by overall market volatility. For example, in October 2007, 50% of all ETFs had an average spread of $0.05/share or less. By October 2008, with the markets in the midst of crisis, only 16% had a spread that small.6 Viewed as a spread percentage, which divides the dollar spread into the share price, 60% of all ETFs had spreads of 0% or 0.10% in October 2007, while only 16% had spreads that narrow in October 2008. Perhaps more troubling for ETF investors during the turmoil was that 62% of all ETFs had spreads wider than 0.50%.7

Liquid, solid or gas

APs play a critical role in the ETF industry. ETF investors, whether they realize it or not, are making a bet not just on the security they purchase, but also on the AP. While the AP (or, in some cases, multiple APs) are really what makes the ETF function, there are no guarantees of active AP trading throughout the day, which can create problems for investors.8 For example, a connection can be made between potential late-session spread expansion and the related AP’s appetite for accepting the potential risk of holding securities overnight.

Additionally, there are AP-related issues that seem to emerge largely at random. One of the more recent examples occurred when Knight Capital Group Inc. came under pressure after a computer glitch last August. As a result, spreads on low-liquidity ETFs where Knight is a dominant market maker soared to 1.5%, up from a previous average of 0.49%.9

Pumping irony

For an investor attracted to ETFs by the idea of simplicity, the solution for avoiding risks — doing potentially extensive research — is perhaps more than a little ironic. Beyond the research, investors should also understand how various non-correlated factors can influence ETF performance.

For example, avoiding buying at a premium and selling at a discount is only part of the equation and involves comparing the ETF’s price versus its underlying assets as well as considering premium/discount relationships historically and by sector to gauge volatility. After that, investors may have other concerns related to particular market volatility and might want to consider utilizing “limit orders” instead of “market orders” for ETF purchases. But again, all of that is only part of an equation that must also consider expense ratios, current and historical spreads, the firm operating the ETF and the ETFs exposure to counterparty risks in various markets and connected to its AP or APs.

And the potential concerns don’t end there.
  • Theoretically, the optimal performance of a basic ETF is to match its underlying index before accounting for expenses Typically, however, tracking errors result in a performance that falls below the index. Tracking errors include at least the management fees, but can also be larger for various reasons. For example, in 2009 ETF tracking errors averaged 1.25%.10 Although the average has come down in recent years, it varies by fund and by sector with higher errors generally in niche funds. For example, in 2011, one niche ETF had a tracking error that shaved 10% off its actual returns.11 In some cases, tracking errors may be a sign of structural issues within the modified portfolio held by an ETF or may suggest liquidity difficulties or other problems. As a result, financial advisors often screen out ETFs with high tracking errors before even considering other factors such as expense ratios.12
  • For investors wanting to utilize an ETF to gain outperformance, there are leveraged ETFs that seek to magnify index returns. However, investors may not understand that the return is measured only against single-day performance. These ETFs, which started to appear in the mid-2000s, are designed only for extremely short-term market bets. Investors need to be aware that while they hope to magnify returns with these ETFs, the risks also are magnified. Prospectuses for leveraged ETFs often note that holding the product for more than one day will result in losses because of daily compounding.13 Depending on the time frame, the losses can actually increase as the underlying index moves higher while the correlation with the index’s overall movement declines.14 Additionally, expense ratios for these products are often high.
  • Some research on active versus passive management has found that a passive approach may have merit for investors who want short-term exposure to a specific asset class.15 However, short-term and aggressive buying and selling of ETFs can magnify risks such as the bid-ask spread issue and the potential for premiums. For the cost structure to be most effective, ETFs would have to be held for longer periods so the costs could be averaged out.
Running with the pack?

Although ETFs have been around for 20 years, their popularity has risen in recent years with investors hoping to find a cheap, simple strategy to keep up with the market. Those who want the best chance of meeting their long-term financial goals, however, should be prepared to do their homework on a wide range of topics to gauge the true “cost.”

Although some investors may prefer the safety of numbers, ETFs certainly cater to the market’s herd mentality that may, depending on the ETF and the market, make investors more vulnerable to sector-specific issues or other problems. For example, because of market weighting, Cisco would have held a 5% position in a Russell 1000 Growth Index tracking fund in 2000, while information technology overall would have been 50% of the fund. Similarly, financials were more than one-third of the Russell 1000 Value Index prior to the 2007 crisis. Although these examples may be extreme cases — and the outcomes would have been historically bad for index-tracking investors — the same effect is always at work on every index. For example, information technology has accounted for at or near 19% of the S&P 500 at year-end for the past three years.

Obviously, the events of 2000 and 2007 were in no way isolated to ETFs. And while ETFs may contain some unique risks, they have the potential to benefit investors who understand those risks and the role an ETF may play in a broader investment strategy. There are sometimes significant differences between passive investing through an index-tracking product of any type, and an active strategy, which has the potential for reallocation and outperformance. Investors may want to keep in mind that, at their core, even the best ETFs are confined to violating one of legendary investor Sir John Templeton’s most famous maxims: “It is impossible to produce a superior performance unless you do something different from the majority.”


1 Investment Company Institute data, January 2013
2 Inefficiencies in the Pricing of Exchange-Traded Funds, NYU Stern School of Business, Antti Petajisto, Aug. 29, 2012
3 Analysis of data at: http://funds.us.reuters.com/US/screener/screener.asp
4 “Junk bond ETF investors rush for the exits,” Reuters, Nov. 15, 2012.
5 Exchange traded funds: When are there too many?, USA Today, Oct. 10, 2012
6 ETF Spreads Widen Substantially, Index Universe.com, Oct. 30, 2008.
7 ETF Spreads Widen Substantially, Index Universe.com, Oct. 30, 2008.
8 Global Financial Stability Report, April 2011
9 Some ETF Spreads Widen as Knight Capital Seeks to Survive, Bloomberg, Aug. 3, 2012
10 IndexUniverse.com, March 9, 2012
11 “Morgan Stanley: ETF Tracking Error Dropping,” IndexUniverse.com, March 9, 2012
12 ETFtrends.com Jan. 14, 2013
13 Beware ‘Leveraged’ ETFs, May 11, 2012
14 Are Leveraged ETFs Just Double-or-Nothing Bets? U.S. News and World Report, July 9, 2012
15 Re-thinking the Active vs. Passive Debate, Goldman Sachs

Past performance is not a guarantee of future results. The opinions expressed in this article are those of the Waddell & Reed Advisor Funds and its fund managers and analysts, and are not meant to predict or project the future performance of any investment product. The opinions are current through March 26, 2013, are subject to change at any time based on market and other current conditions, and no forecasts can be guaranteed..

Investors should consider the investment objectives, risks, charges and expenses of a fund carefully before investing. For a prospectus, or if available, a summary prospectus, containing this and other information for any of the Waddell & Reed Advisors Funds, call your financial advisor or visit www.waddell.com. Please read the prospectus or summary prospectus carefully before investing.

Investment return and principal value will fluctuate, and it is possible to lose money by investing. Past performance is not a guarantee of future results.

Financial Advisor Opportunities
Corporate Careers