Volatile market conditions mean buyside traders must take a hard look at their brokers’ algorithmic trading practices.
That’s the message from Deutsche Bank Securities in a just-released paper analyzing the effects of the recent bout of high volatility on key trading metrics.
Because the performance of trading algorithms is based on historical data, Deutsche Bank argues, buyside traders need to be sure the algorithms they use also take into account current conditions.
“Additional controls via conditional order instructions are necessary to ensure the order is completed within the original objectives as real-time data changes,” the report, authored by managing director Rob Flatley, states.
The report also emphasizes that traders must make sure the algorithms they are using are able to reach as many pools of liquidity as possible via smart order routing.
Behind Deutsche Bank’s recommendations are changes in the character of the stock market brought on by volatility caused by uncertainty over economic and financial issues.
Volatility, as measured by the Chicago Board Options Exchange’s VIX index, has surged since August 2007. It jumped from about 15 to 30 in August and has stayed in the mid-20s since.
Volume has jumped as well, with trading in S&P 500 names hitting an average of 4.7 billion shares in January. That compares to less than 3 billion shares in January 2007.
The trading surge and accompanying price swings have impacted such trading metrics as spreads, top-of-book quoted size, average trade size, and velocity.
Spreads in S&P 500 names, for example, were over eight basis points in August 2007, November 2007 and January 2008, the three most volatile months covered by the study. That’s after averaging between four and five basis points in the preceding months of 2007.
Top-of-book quoted size in S&P 500 names was between 30 percent and 40 percent lower in the same three months than it was in the first quarter of 2007, the report noted.
Average trade size plummeted, the report said. Among large cap stocks, it dropped 25 percent from the start of 2007.
Finally, velocity, or the number of fills per second, also increased over the course of the year.
“You have these whipsaw events,” Flatley told Traders Magazine recently. “Systems are data-driven. Data-driven systems are great, but they work on averages. What happens on a non-average day?”
