The growing body of HFT research: time to put old arguments to bed?

30 March 2017


Over the last few years the great and the good of the world’s regulatory authorities that govern financial markets have studied the impacts of the electronification of the markets, and the role played by high frequency trading (HFT).

Each has used a particular methodology and data set, as befits their jurisdiction, to examine a particular element of HFT.  One by one, the myths that underpin the criticism of HFT have been assessed and dispelled. To add to this, we are only in the early part of 2017 and we are already seeing more reports de-bunking the myths around HFT.

Does HFT only create ‘ghost liquidity’, asks the Dutch AFM?  No, finds their report.

Does this liquidity disappear in times of stress? No, finds BIS.

Are high-frequency traders anticipating the order flow? The FCA finds no evidence of this.

Were HFT firms solely responsible for the sterling flash crash in 2016? Again, BIS says no.

Much of this research has gone under the radar, especially compared to the often vocal and well-publicised criticism that propagated these myths in the first place. So it’s easy to overlook the number of reports by respected, independent bodies that have now examined the topic, from a position of expertise and access to the data.

We welcome the recognition in these studies that markets and end users benefit from increased use of innovative technology, in line with the experience with innovation in other industries.

Here’s a summary of the recent reports:

The sterling ‘flash event‘ of 7 October 2016 (BIS), January 2017

This report by the Bank for International Settlements (BIS) examines the so-called Sterling Flash Crash of October 2016. Drawing on analysis from the Bank of England, as well as information from the BIS Markets Committee, the analysis points to ‘a confluence of factors catalysing the move, rather than to a single clear driver.’ The report also finds that the time of day was a significant factor in increasing the vulnerability of the Sterling Foreign Exchange market. It notes that events such as this have been short lived and have not significantly impacted financial stability.

Gauging market dynamics using trade repository data:  the case of the Swiss franc de-pegging (Bank of England), January 2017

This report by the Bank of England looks at the market impacts of the decision by the Swiss National Bank to discontinue the Swiss franc’s floor of 1.20 Swiss francs per euro. The report examines the effects of the move on the market in the short and long term. In the short term, it found that the removal of the floor led to extreme price moves. It found evidence that the ‘rapid intraday price fluctuation was associated with poor underlying market liquidity conditions.’ In the longer term, it found a reduced level of liquidity, associated with an increased level of market fragmentation and higher market volatility.

Working Paper Series: The stock market effects of a securities transaction tax: quasi-experimental evidence from Italy (ECB), August 2016

This working paper by the ECB focuses on the impact of the Financial Transaction Tax (FTT)/Security Transaction Tax (STT) on Italian markets. The paper finds the FTT in Italy has negatively impacted the markets. The study explains that ‘overall the introduction of the STT induced a reduction in liquidity for the stocks hit by the reform. We also find some evidence that volatility of the "treated" stocks increased.’ The study also finds that the introduction of the FTT has led to a wider bid-ask spread. It notes that there was insufficient empirical data to make a full assessment of the effects of the tax on OTC markets.

A case analysis of critiques on high-frequency trading (AFM), June 2016

In this report the AFM looked at two often expressed critiques of HFT: that HFT is accused of ‘ghost liquidity’ and that HFT strategies can make riskless profits at the expense of investors. For the first part of the report, the AFM found that ‘underlying trading patterns are a logical consequence of the application of market making strategies in a fragmented marketplace.’ For the second part of its report, the AFM, having analysed large orders at several venues, could not find any evidence to support that HFT firms make unfair profits at the expense of investors. They say ‘we did not find any evidence of HFT executing the liquidity detection strategy in our case studies.’

The Economics of High-Frequency Trading: Taking Stock (Albert J. Menkveld), June 2016

In this report, Menkveld states that high frequency traders ‘helped us migrate quickly to electronic trading which, in turn, yielded lower transaction cost and more volume’ and that being well-informed they become market makers. He also concludes that HFT firms provide venue competition with more innovation and lower trading costs. Overall, Menkveld says that ‘bottom line: I believe economic benefits outweigh costs.’ He also outlines what he believes are the three main benefits of HFT, which are ‘market making, venue competition, more trading opportunities.’

Order duplication and liquidity measurement in EU equity markets (ESMA), June 2016

This report focuses on the change in the trading landscape of equity markets over the last decade. The study used unique data collected by ESMA and covers a sample of 100 stocks on 12 trading venues in nine EU countries. The study found that order duplication, used by traders to ensure execution across multiple trading venues, contributes positively to liquidity. It concludes that overall multi-venue trading has increased the liquidity in EU equity markets.

BIS Working Papers, No. 563. Who supplies liquidity, how and when? (BIS), May 2016

This paper looks at who provides liquidity in the limit order books, by using data from Euronext and the AMF to observe how traders and the market connect. Their research found that liquidity supply by proprietary traders ‘does not evaporate when it is most needed.’ It found that proprietary traders, fast or slow, provide liquidity with contrarian marketable orders, thus helping the market to absorb shocks, even during a crisis. It also concludes that market reforms may have negative consequences, stating that: ‘new banking regulations, making it more difficult and costly for banks to engage in proprietary trading, might also reduce market liquidity.’

Occasional Paper No. 16: Are high-frequency traders anticipating the order flow? Cross-venue evidence from the UK market (FCA), April 2016

This FCA paper investigates claims that firms using HFT exploit their speed advantage and anticipate when orders arrive at other, slower, trading venues. It also looks at whether HFT firms can anticipate the order flow over longer time frames. For the first part of the report, the FCA can find no evidence that HFT anticipates near-simultaneous orders. They found that ‘HFTs are not systematically observing an order on one venue, predicting that a similar order will be arriving at another venue.’ While the FCA found patterns consistent with HFT being able to anticipate the order flow over a long period, no concrete evidence supports this. The report states that HFT firms may simply be ‘reacting more quickly to news and other public information.’

Report on algorithm trading on NASDAQ Copenhagen (FSA, Denmark), February 2016

Algorithms have improved liquidity in securities trading at Nasdaq Copenhagen with no evidence of increased price volatility or flash crashes caused by erroneous programming, according to the Danish FSA in its report. ‘The widespread use of computerized trading looks, according to our research, to have resulted in improved liquidity in the market. There is no evidence of greater risk for example, from greater fluctuations in prices as a result of algorithmic trading.’  Algorithms were found to account for 50% of all trading with HFT amounting to 15% of total volume. The FSA found no evidence of market abuse by algorithm-based trading strategies.

Working Paper No. 523 Interactions among high-frequency traders (Bank of England), February 2015

This paper examines the interactions between different firms using HFT and the impact of such interactions on price discovery. The study finds that correlated trading by HFT firms is associated with permanent price impact but that it is information based and that firms are ‘reacting simultaneously and quickly to new information as it arrives at the market place, which makes prices more efficient.’ The paper also found that correlated trading by HFT firms ‘does not appear to contribute to undue price pressure and price dislocations on a systematic basis in the UK equity market.’ The paper also states that commonality among the algorithmic trading strategies in the foreign exchange market does not create ‘price pressures and excess volatility that would be detrimental to market quality.’

  • Blog