The history of high frequency trading

April 11, 2016
Editor(s): Alex Kent
Writer(s): James Clements, Fiona Mei Ling Wong, Sarinie Ning, Sasank Pazhannur

Stock and securities trading has evolved immensely over the course of history from the time the first stock exchange was opened in Amsterdam in 1602. Whilst the system has its grassroots in the age old barter system, and has come to exist as a highly complex and integrated worldwide phenomenon.

However, only in a recent history we have seen another revolution in the market, as commodity trading on Wall St has evolved into a more modernized, digitalised and efficient method known as high frequency trading (HFT).

HFT is a highly automated method of trading that utilizes complex algorithms to enable firms to trade large volumes of stocks within milliseconds. Compared to traditional trading, it ignores long-term valuation techniques and instead searches for arbitrage opportunities, pricing inefficiencies and spot trends using mathematical formulas.
The introduction of HFT, which took place in the early 2000’s, was brought about due to advancements in computational technology, enabling these high-speed and high-volume trades. The idea of earning healthy profits rapidly attracted investors across the globe, and within a short interlude, HFT had increased to account for 10% of all trades. To exemplify the rising popularity of HFT, the graph below illustrates HFT as a percentage of equity turnovers.

Indeed, HFT had become so popular that in 2010 equity holdings from computer-based transactions were valued up to $4.1 billion. As a result, events such as the ‘Flash Crash of 2010’ have extensive impacts on the market. This has sparked global-wide debate surrounding the security of the perceived risky HFT system.

Methods of HFT

Competition between highly aggressive HFT firms has turned short term trading into a rat race, with the winner being who can employ computer algorithms the fastest and most accurately in order to capitalise on infinitesimally small arbitrage opportunities.

This begs the question how is this speed game implemented. Unsurprisingly, firms are somewhat secretive about the sophisticated algorithms they employ. Most of these strategies rely on the rapid speed of operating systems, often with trade taking place within the microsecond category.

The list of strategies used for HFT is extensive, but some common methods include:

  • Market making – Firms earn profits by placing a high volume of orders to both sell and buy the same stock, often with a smaller bid-ask spread
  • Information arbitrage – Computers identify emerging trends in media publications, such as Bloomberg news, and quickly respond to the information by trading relevant stock.
  • Statistical arbitrage – Highly quantitative algorithms seek out short-term discrepancies in prices between securities, and capitalise on these inconsistencies
  • Momentum ignition – More controversially, many HFT firms also make profit through prompting and predicting the actions of other algorithms. Computers artificially create price movements, inciting actions by other algorithms and hence enabling the initial instigator of the price movement to capitalise on this expected response and the resulting price movements

This list of strategies merely scratches the surface of methods HFT companies use to earn profits with their short holding period portfolios.

Risks and controversies of HFT

Following the Flash Crash of 2010 in the United States, the risks and controversies of HFT have received widespread public scrutiny. In this particular event, the chaos arose when the Dow Jones Index plummeted 1000 points in a matter of minutes, causing billions of dollars to be wiped from the exchange. One of the key causes of the event has been identified as a chain reaction caused by a HFT algorithm.

So, what are the risks that come with engaging in HFT? According to Picardo (2016), one of the biggest risks is the systemic risk; risk that affects the entire market. The reason systemic risk is amplified through HFT is due to the strong linkages between financial markets resulting in shocks being rapidly transmitted from one market to the next, known as the ‘ripple effect’. As the markets and asset classes are highly integrated in today’s day and age, a meltdown in either one can ripple across other markets and asset classes in a chain reaction.

Another risk associated with HFT is errant algorithms. Faulty algorithms can result in significant losses within a short period as HFT takes place in milliseconds. An illustration of this was the $440 million loss suffered by market-making firm Knight Capital, in August 2012, in a mere 45 minutes due to a defective algorithm which eliminated the bid-ask spread. Rival traders took advantage of this situation and by the time Knight salvaged the problem, the firm had been pushed close to bankruptcy.

A more controversial implication surrounding HFT is market manipulation. In addition to the aforementioned ‘momentum ignition’, another contentious technique is ‘spoofing’, which occurs when a participant places a fake competing buy order to encourage the original participant to increase its offer price. The participant then cancels the order and places a new sell order, with an increased offer price, generating a tidy profit. This illegal activity misleads the supply and demand of financial instruments in the market.

Protective measures have been implemented to combat these risks. In 2014, an additional safety level called the “kill switch” was introduced by the ‘Nasdaq OMX Group’ for its member firms to halt trading if a specified risk exposure level is breached. Other measures regarding transparency and accountability have also been instigated.

The positive impact of HFT on markets

It must be noted that the impact of HFT on markets is not completely negative. HFT promotes a highly competitive environment compared to traditional market structures, with studies showing it enhances liquidity in the market and enhances market efficiency by reducing bid-ask spreads.

In the past 20 years, bid-ask spreads have dropped dramatically, with one reason being cited as HFT. HFT enables traders to profit from a high volume of trades with relatively smaller spreads, reducing the friction in selling assets market wide.

Similarly, HFT plays a role in price discovery, decreasing excessive price volatility. Several academic studies have focused on documenting the improvement of price discovery through HFT. According to a study conducted by Brogaard, Hendershott and Riordan (2014), HFT improves the efficiency of pricing by trading in the direction of permanent price change and in the opposite direction of transitory pricing errors. In essence, its informational advantage in marketable orders is sufficient to overcome the bid-ask spread and trading fees to generate positive trading revenues. By leveraging on electronic trading tools and high frequency financial data, this algorithmic trading method is able to identify the abnormality in price and make arbitrage possible. Accordingly, high frequency traders will deploy arbitrage strategies to capitalize on the irrational price differences and eventually push prices back towards equilibrium.

In summary, HFT is undoubtedly a tool that has become an integral part of the stock market. Nevertheless, scrutiny regarding its disruption to traditional markets structures and methods is never far from the spotlight.

Editor: Alex Kent

Writers: James Clements, Fiona Mei Ling Wong, Sarinie Ning, Sasank Pazhannur

The CAINZ Digest is published by CAINZ, a student society affiliated with the Faculty of Business at the University of Melbourne. Opinions published are not necessarily those of the publishers, printers or editors. CAINZ and the University of Melbourne do not accept any responsibility for the accuracy of information contained in the publication.

Meet our authors:

Alex Kent
James Clements
Fiona Mei Ling Wong

This author has not left any details

Sarinie Ning
Sasank Pazhannur