top of page

In the blink of an eye


Current estimates show that around 50% of the total volume observed in U.S. capital markets derives from high-frequency trading (HFT) transactions. This figure may appear surprising at first glance, but it is perfectly in line with the nature of these new trading platforms, which use extremely powerful computers to place a huge number of orders at an unprecedented speed. Moreover, the recent listing of Virtu Investments ­– which has only been in the red once out of 1238 trading days – is a case in point for the insurmountable rise of HFT. The monumental status of the HFT behemoth has attracted considerable attention by regulators and the media, but the jury is still out on its impact on the market. Unfortunately, the contention that surrounds HFT has led to the proliferation of various misconceptions.

A set of distinguishing features can be identified to unveil the essence of HFT. They usually include the following: the employment of supercomputers to manage (execute, route and generate) orders, the contracting of co-location services (i.e. renting server space in the proximity of an exchange in order to obtain market data faster), the maintenance of insignificant positions (i.e. low capital at risk) on a vast array of individual securities for a short period of time, the perpetration of fleet orders (i.e. cancelling orders before they are executed), and the absence of unhedged overnight positions.


Before diving into the murky waters of HFT, a few rudimentary concepts need to be clarified. HFT is a subset of automated trading that is carried out at an extremely high speed. Meanwhile, automated trading simply refers to the complete mechanisation of trading processes. This involves monitoring risk, financial modelling and algorithmic trading. More specifically, algorithmic trading (algo trading) comprises a predefined set of trade execution techniques that attempt to reduce risk while respecting certain time constraints. This usually involves slicing up larger trades to ascertain price stability while reducing market impact (i.e. information dispersion).

Many HFT firms and institutional investors go a step further by resorting to dark pools. By trading on these private exchanges investors remain anonymous and avoid revealing their trading strategies. This considerable advantage vis-à-vis ordinary trading platforms comes into play with block trades (i.e. large trades). Since quotes are not publicly displayed, other market participants cannot exploit information leakages and large sales do not lead to coincident declines in the price of securities. On the other hand, an inevitable drawback is a lack of transparency that enables conflicts of interest to develop. For example, operators of dark pools can choose which orders to internalise (i.e. fill with their own inventory of securities) and which to redirect to an exchange. This enables them to pick the most favourable spreads and offload toxic orders. Moreover, some brokers grant special access to proprietary trading firms in return for a fee to the detriment of other pool clients.


Clearly, HFT has become a key component of the financial ecosystem but its remarkable evolution over the the past fifteen years only took on a dominant role following the liquidity crisis instigated by Lehman Brothers collapse in 2008. The aftermath of the crisis led major exchanges to start offering incentives to firms in order to increase competition, so-called volume, and consequently provide liquidity to the market. Theoretically, the use of high-powered computers constitutes an attempt to improve the overall welfare of market participants. However, a concurrent shift in the scope of activities carried out by stock exchanges has occurred. They have evolved from their not-for-profit nature into decentralized profit-taking venues.

There are several ways in which high-frequency traders attempt to exploit market opportunities. Primarily, they do so by observing institutional investors and anticipating their moves on the market. Computer programs can interpret big players’ behaviour and capitalise on various trading techniques. For example, anticipating that an investor is about to buy a security, an algorithmic trader will go long, outpace the investor and close the position immediately after the investor’s transaction has been carried out (i.e. micro front running). Another source of profit is statistical arbitrage. With the right algorithm, high-frequency traders take advantage of minuscule price differences between exchanges. Buying the cheaper stock and selling the marginally higher one ­­–­ in the blink of an eye ­­– allows traders to make an almost risk-free profit.


Citadel LLC’s performance in recent years is indicative of HFT’s effect on the market. Between 2008 and 2012, Citadel earned a 31% return by engaging solely in HFT stock and futures transactions. In stark contrast, the S&P 500 and hedge funds’ returns plummeted by 37% and 19% respectively. In 2009 Citadel started using statistical arbitrage, thereby reaching peak revenues of $4.9 billion. The surge in earnings derived from HFT elicited a spike in new entrants. As a direct result, Citadel’s total revenues dropped to $810 million by 2012 according to Rosenblatt Securities Inc., a New York-based brokerage firm. The aforementioned dynamics are represented in Michael Lewis’ book Flash Boys, published in March 2014. According to his observations, firms have started diversifying their revenue stream by introducing non-HFT strategies to give consistency to their returns. Nevertheless, a colossal $23 trillion are still tilted in favour of speed traders who take advantage of slower investors by obtaining order information earlier.

In the influential paper The Dark Side of Trading, Ilia D. Dichev, an Accounting Professor at Emory University, points out that HFT distorts the work of traditional investors who focus on financial statement analysis. Moreover, retail investors who trade based on conventional market feeds now have to face the fierce competition of supercomputers, which create idiosyncratic market patterns based on speculative strategies. Therefore, the large volume of transactions generated by HFT does not necessarily reflect the fundamental value of stock prices (i.e. the real nature of firms), thereby increasing market volatility. Another complication that arises as a result of the manipulative strategies pursued by high-speed trading firms, involves placing a vast number of orders that are cancelled soon after submission (i.e. fleeting orders). This creates the false appearance of trading activity and severely distorts market trends.


In September 2012, Haim Bodek, founder of the HFT outfit Trading Machines, blew the whistle on Direct Edge. ­­The computerized exchange had allegedly been granting preferential access to his direct competitors. After Trading Machines suddenly began losing out on large orders, Haim Bodek suspected that unfair practices were taking place. His doubts turned out to be justified after a Direct Edge employee revealed to him that Hide Not Slide orders were secretly being executed before plain vanilla limit orders. Finally, the SEC had proof that certain market orders were hurting unwary traders. Meanwhile, investigators of the U.S. Commodity Futures Trading Commission (CFTC) uncovered that proprietary trading firms were simultaneously taking part in transactions as buyers and sellers, thereby distorting the futures market and committing what under U.S. regulation is known as “wash trading”.

Moreover, in 2013, the Federal Bureau of Investigation (FBI) launched an enquiry to establish whether HFT firms were engaging in insider trading. It was investigated whether certain large orders were carried out to hide smaller ones based on private information. In addition, U.S. General Attorney Eric Holder and New York General Attorney Eric Schneiderman analysed low latency data feeds that allegedly gave high-speed traders an unfair advantage. Despite the complexity of the underlying subject matter, a team composed of the FBI, the Securities and Exchange Commission (SEC) and the CFTC found evidence of preferential treatment. More recently, several HFT brokers were charged with abusing their fiduciary duties by trading on their own account using information derived from clients’ orders. It was also suspected that after-hours information was used outperform the market the following day.


Clearly, HFT is a double-edged sword. The benefits of improved liquidity are offset by the predatory practices of sophisticated trading firms who reside in the esoteric realms of quantitative finance. Regulatory institutions are incapable of keeping up with the intricacy of current developments. Charges that primarily lead to insignificant fines do not solve the problem but rather exacerbate it by undermining the authority of law enforcers. The inextricable consequence of lax regulatory measures is a moral hazard problem of gargantuan proportions. An earnest approach to reform would necessitate a complete overhaul of the high-speed trading landscape that involves both regulators and insiders. Otherwise Pandora’s box of HFT-related offences will keep on wreaking havoc.




Recent Posts
Search By Tags
Follow Us
  • Facebook Basic Square
  • Twitter Basic Square
  • Google+ Basic Square
bottom of page