Some high-frequency trading (HFT) techniques can exacerbate price swings and certain momentum strategies may have tried to profit from the recent market turmoil and may even have caused some volatility in order to extract information.
“There has been much research that shows that some HFT strategies tend to cause prices to overshoot their fundamentals and therefore cause volatility. It is also especially important to differentiate between HFT suppressing volatility and HFT not participating in volatile markets. Just because HFT may not participate in volatile markets doesn't mean that they are responsible for stabilising prices,” said Will Psomadelis, head of trading, Australia, at Schroder Investment Management.
Recent papers – such as The Effect of High-Frequency Trading on Stock Volatility and Price Discovery by X Frank Zhang of Yale University's School of Management – have supported some of the concerns of market participants about HFT, but the debate remains a live one.
Psomadelis says HFT cannot be blamed solely for the volatility. “HFT takes many forms and to point the finger, either in a positive or negative way, at such a broad trading genre does not contribute to the global debate about their impact on markets.”
Many commentators have claimed that increased activity by HFT firms may have contributed to volatility, but recent empirical evidence from Australia-based research consultancy Capital Markets Cooperative Research Centre (CMCRC) suggests the opposite is more likely.
Based on recent studies of the Australian, French and Singaporean equities markets, Alex Frino, CEO, CMCRC, said HFT tends to stabilise markets during periods of market turmoil. There's very little HFT activity in Singapore's equities markets, and CMCRC has not found any significant relationship between HFT activity and price volatility for Singapore's derivative markets.
According to some estimates, HFT firms are on one or other side of 60% of equities trades executed on the Australian Securities Exchange.
“In Australia, we find a negative relationship between HFT and price volatility for equity markets. This means rather than exacerbating price moves, they have a stabilising effect on prices,” Frino said. In some cases, precisely because they ignore the fundamentals, HFT strategies can apply downward pressure to market overreaction.
“We do not find, for any of the markets that we've looked at, a positive relationship between HFT activity and price volatility. On this basis, we do not find credible the recent assertions that HFT activity is contributing to market volatility,” he noted.
Bryce Kelly, head of product strategy, Asia, at Instinet, takes a similar view, “HFT are market-makers and volatility sellers, so in weeks like w/c 8 August they tend to see a lot of opportunity and likely ramp up trading activity. But it's probably a stretch to say that they were a cause of the volatility.”
Volatility indexes across the region has spiked since early August when Standard & Poor's downgraded the US' credit rating. On 5 August, the Australian VIX spiked 26% to 31.86 while Hong Kong's HSI Volatility Index closed at 33.87, an increase of 12.19 points or 56.23% from the previous day's close. The HSI Volatility Index stood at 38.63 on 23 August.
Recent weeks have also seen the return of interest in trading volatility as an asset class. Guillaume Derville, head of derivatives strategy marketing for Asia Pacific, global equities and commodity derivatives at BNP Paribas, said, “Some institutional investors have taken advantage of this spike in volatility and are using variance swaps to get long exposure to volatility.”
In Europe, the extreme volatility has led to some buy-side traders abandoning algorithms altogether through fear of being gamed by HFT firms.
But Psomadelis says algorithm usage should not determined by volatility. “Rather, it is part of the strategy mix (along with block dark pools and natural blocks) and is ultimately guided by the level of natural liquidity trading in the upstairs market vs the lit market. Natural-to-natural block trades are the best way to reduce market impact hence why our trading strategy remained unchanged in the volatility we saw in the equity markets a couple of weeks ago.
“Whilst we may have use a different mix of algorithms to interact with various types of liquidity, market volatility was not a factor in deciding how much electronic trading we undertook. Algorithm use is only part of the strategy mix used in an attempt to interact with natural liquidity and should not therefore be confined to a target percentage.”
Kelly noted that that special pricing analytics are built into Instinet's suite of algorithms to specifically manage volatility. “We’re constantly analyzing what’s going on in the market and using that to inform trading decisions.”
Instinet had a record week in its overall Asian regional volumes, as well as in orders directed into Nighthawk, its dark aggregation algo. “We saw a lot of demand to maximize liquidity while minimising footprint,” Kelly noted.
For the week of August 8-12, 2011, the firm traded a record US$2.7 billion in the non-Japan Asia region, and over US$148 million across Asia in 17 alternative destinations traded via Nighthawk.
Author: Jill Wong