One of the main advertised benefits of ETFs is the instant access to market liquidity that the words “exchange-traded” promise. By contrast, the traditional mutual fund model offers investors only a single daily point of entry and exit. But is the option for ETF investors to execute trades intraday a real benefit?
Australia’s financial regulator, the Securities and Investments Commission (ASIC), isn’t sure, particularly when it comes to the retail client. In a recently released set of guidelines on ETFs, ASIC defines price errors, or “gapping”, as a key risk for the general public.
“In some cases ASIC has found examples of ETF prices quoted by online stockbrokers that are significantly above or below the value of the assets that the ETF holds. The risk is that you might pay far more than the ETF’s assets are worth, or sell ETFs at a price far below the value of their assets,” says the regulator.
“Before placing an ETF order, check the price you’re quoted matches what the ETF issuer says its assets are worth (the ‘net asset value’, or NAV),” advises ASIC. “Some ETF issuers’ websites regularly update their estimated NAVs.”
The risk of executing an order to buy or sell an exchange-traded fund at an uneconomic price was highlighted by last year’s flash crash. During the flash crash many ETFs, and a far greater proportion of ETFs than stocks, suffered large-scale intraday losses. Many investor orders to sell ETFs were executed against so-called “stub quotes” of fractions of a dollar and were subsequently cancelled. Following the event, US regulators introduced new “circuit-breakers” for the largest US companies and for 344 ETFs, imposing a five minute trading pause when a stock or ETF moves by 10 percent. The circuit-breakers have since been made permanent and extended to all US-listed securities.
The flash crash was a single, record-breaking market-wide malfunction, which many experts have since attributed to rapid technological change, regulatory inconsistencies and the complexities of market structure. But even in non-crisis periods there are potentially very large differences in the day-to-day costs of trading ETFs. During periods of market volatility these can become even more pronounced.
For example, according to the London Stock Exchange, the average cost of a round-trip trade in an ETF was 2.46 percent in the week ending August 12. The LSE measures spreads on a time-weighted average basis, using individual bid and ask (“tick”) prices from on the exchange’s order book. That’s up from 0.74 percent three weeks earlier, reflecting the impact of August’s market volatility. By this measure, buying and selling the average ETF cost over six times more than the typical fund’s annual expense ratio (measured by BlackRock recently as 0.37 percent).
But disguised within the measure of average dealing spreads there’s a remarkable range of ETF trading costs.
At one end of the scale, the LSE’s most traded ETF, iShares’ FTSE 100 fund (LSE: ISF), recorded an average time-weighted spread of 9 basis points in the week ending August 12, up from 3 basis points for the week ending 22 July. ETF Securities’ Physical Gold ETC (PHAU), which is typically the second most traded tracker on the exchange, saw a rise in its time-weighted spread from 6 to 8 basis points over the same period.
At the other end of the scale, however, during the week ending August 12 the LSE recorded time-weighted average bid-offer spreads of over 10 percent in 26 of the 623 ETFs listed on the exchange. Twelve funds had an average spread exceeding 20 percent, and three had spreads of over 50 percent.
According to iShares, the issuer of 10 of the ETFs with recorded time-weighted average spreads of over 10 percent during that week, this particular measure of ETF tradability may give a misleadingly pessimistic view of actual market liquidity. iShares pointed out to IndexUniverse.eu that, during the week of August 12, high market volatility meant that there were periods when ETF market makers stopped quoting altogether. In the absence of market maker quotes, any time-weighted average spread measure would be based either on limit orders that might be far away from an ETF’s true value, or even on bids at a zero price.