The growing use of new types of computer models that react more quickly to the latest moves in prices is exacerbating the historically high levels of volatility in equity markets, according to traders.
Such “aggressive” algorithms that are based on the most up-to-date price data – often for just the previous hour or day on a particular stock – have replaced models using longer term valuation methods and historical data going back 10 years or more.
Sudden recent moves in prices can therefore prompt computer-based traders to quickly move into or out of a particular market.
This is fuelling the sudden swings in direction and heightened volatility that have become the norm in the US equity markets in recent times.
The CBOE’s Vix index of market volatility, known as Wall Street’s “fear gauge”, hit highs of close to 90 in October and has since fallen back to about 60 but still indicates heightened levels of distress.
In the past, a large proportion of algorithmic trading was accomplished using a simple model know as the Volume Weighted Average Price algorithm that uses historical data to calculate the average price at which a stock trades over a particular time period.
The theory is that if the price of a particular buy trade is lower than the VWAP, it is a good trade and will automatically go through and if it is higher, it is a bad trade and the algorithm will “move on”.
Goldman Sachs confirmed in a recent study that algorithmic trading clients had shifted away from the relatively passive VWAP algorithm to more aggressive models and within those models, were choosing more aggressive settings.
Goldman also pointed out that, amid the market turmoil, the use of algorithmic trading had remained steady, with algorithms making up 15 per cent of the total value executed by clients through Goldman Sachs in September compared with 16 per cent in May.
John Bohan, managing director of global execution services at BNP Paribas in New York, said: “With 10 per cent intraday swings in the S&P 500 index experienced recently, investors and traders are becoming less concerned with reducing the market impact of their activities and have drawn back from using VWAP algorithms.”
He added that many traders are now seeking out liquidity at all costs, finding a particular price, filling an order and then quickly moving on to the next price level.
“This has a knock-on effect on the market and is adding to the volatility in what would most accurately be described as a snowball effect,” he said.
Mark Palmer, chief executive officer of Streambase, an algorithmic trading software provider, said: “Amid the market turmoil, what you have to remember is that the market structure is shifting fairly rapidly. There are now more venues to trade and more volatility at the same time. At times of greatest volatility, algorithms tend to ‘whiplash’ and feed off one another.”
