A new report from agency broker ITG has highlighted a greater use of liquidity-seeking algorithms for US exchange-traded funds (ETFs) but notes that institutional trading of such instruments is surprisingly low.
According to the report, ‘Institutional trading in exchange-traded funds’, 53% of buy-side trading in ETFs was executed via algos, similar to the proportion of equities trading using algos in the sample analysed.
Last year, schedule-based strategies accounted for 35% of algo-based executions down from 51% in 2010. Liquidity seeking strategies accounted for 24% of ETF algos, compared to 19% for implementation shortfall, 13% for DMA and 7% for dark strategies.
The cost of trading using a liquidity-seeking algo was 0.83 bps, less than half the 1.91 bps cost observed when using a scheduled-based strategy, according to the ITG research.
The report also compared the spreads and volatility of two of the most liquid ETFs, the SPDR S&P 500 and the iShares Russell 2000 index ETF, and their underlying constituents. It found narrower spreads and lower volatility in ETFs compared to their stock equivalents – 1 bps spread and 17% average 60-day historical volatility for the SPDR ETF, compared to a 5 bps spread and 29% average volatility for S&P 500 stocks.
Trading costs based on an implementation shortfall benchmark were also examined, with ETF fees ranging from 1-3 bps for an IS benchmark, compared to 6-60 bps for non-ETFs.
“Trading costs for ETFs were lower due to their relatively better liquidity characteristics, most particularly in the top ten ETFs by volume,” read the report.
Despite the perceived their perceived benefits, only 5% of institutional trading activity was ETF-based last year, compared to 4% in 2010, although 31% of institutional investors in ITG’s survey traded over 5%.
“The numbers suggest that ETFs are cheaper to trade by a wide margin, relative to single stocks,” read the report. “We can rule out simple things like order size as a driver of such results, but issues surrounding liquidity, risk, and associated trading costs have been shown to be quite complex. In the context of practical ETF trading implementations, this area deserves further scrutiny.”