High Frequency Trading (1): Empirical Assessment
By Tabraiz Lodhi
The Nature of HFT
High-frequency trading (HFT) is a type of algorithm-based trading found in financial markets. Financial institutions such as investment banks and hedge funds often have to execute a large number of trades at once. To facilitate this process, they use automated, pre-programed trading instructions, one type of which is high-frequency trading. The distinctive features of HFT are its sophisticated algorithms, high turnover rates and order-to-trade ratios, as well as its short-term investment horizons. High frequency trades involve moving in and out of trading positions at super high volumes and speeds, often conducting hundreds of trades in fractions of a second. These features make HFT a particular unique kind of algorithmic trading.
The use of HFT by financial institutions has accelerated rapidly since its inception. In the mid-1990s, it accounted for less than 3% of the equities market — by 2010 (the year of the Flash Crash), it accounted for over 70% of dollar trading volume. (Zhang, 2010) Today, some industry experts estimate that alogrithmic trading practices account for nearly 90% of trading volume in the equities market, and a comparable volume in the futures market. (Cheng, 2017) Even though the share of volume attributed to HFT has fallen in equities market, in the futures market the use of HFT continues to be prevalent. One reason for this rapid growth is the increasing sophistication of technologies that drive HFT. These include, inter alia, computing power, algorithmic efficiency, and data gathering methods. Given that HFT is prevalent in contemporary financial markets, it is worth exploring how this practice is conducted and relevant government regulations.
Regulations and Practice
HFT is subject to regulatory scrutiny. In the United States, the Financial Industry Regulatory Authority’s (FINRA’s) Rule 3110, is the regulatory regime for HFT. (FINRA, 2014) This rule does not regulate algorithmic trading practices directly. Rather, it establishes general guidelines for supervision so firms may ensure their algorithm-based trading practices are in line with other FINRA regulations. One such regulation is FINRA Rule 5210 regarding the publication of financial transactions and quotations. This rule prevents the practice of self-trade resulting from orders originating from a single algorithm. Another is Rule 2010, which establishes standards of commercial honor. The rule requires firms “to observe high standards of commercial honor and just and equitable principles of trade.” (FINRA, 2015) These rules are not stringent — they mostly provide guidelines and recommendations on the use of HFT rather than any strong regulations.
A more robust regulatory regime than FINRA’s is found in Europe. MiFIR — the Markets in Financial Instruments Regulation — imposes a strict set of organizational requirements on investment firms and trading venues vis-à-vis HFT. These include, inter alia:
- Acquiring authorization to continue to use high-frequency trading techniques;
- Storing time-sequenced records of algorithmic trading systems and trading algorithms for the past five years, and opening these up to government monitors;
- Implementing risk control mechanisms to ensure trading systems are resilient and have enough capacity to prevent sending erroneous orders or contributing to a disorderly market. (Norton Rose Fulbright, 2014)
These provisions are meant to curtail widespread market abuse from specialised investors sending out orders and withdrawing them immediately after using HFT methods. Such practices contribute to market instability and system volatility, and were thought to be in violation of bona fide trading practices. MiFIR’s new regulations have been received favourably by some. (Weller & Bruno, 2018) However, there is still room for improvement. While the transparency requirements under MiFIR and the supervision requirements under FINRA’s rule 5210 are a welcome first step, as Weller & Bruno point out “these requirements alone cannot increase the financial market resilience to shock caused by misuse of high frequency trading practices.” The market needs more direct government regulation of abusive uses of HFT.
Advantages of HFT
- Lower transaction costs
Since the introduction of HFT, transaction costs for retail and institutional investors have decreased by over 50%. (Menkveld, 2016) High-frequency and algorithmic trading are much more cost-effective than traditional trading practices because they eliminate the need for a middle-man. The role traditionally taken on by stock brokers is now being replaced by computers. While the initial capital investment for such technology is steep, the long-term costs are much lower, as there is no need to pay middle men a salary or deal with documentation and associated paperwork. The trend in financial markets then is similar to the trend in many other industries — technology is rapidly replacing certain jobs because of its comparative efficiency and cost-effectiveness.
- Increased liquidity in financial markets
For financial markets to function properly, the presence of so-called ‘market makers’ is essential. These are investors who provide liquidity in financial markets by buying and selling when no one else will. This liquidity is essential for other investors to have confidence in investing. If an investor knows there are people willing to buy his stock if he chooses to sell it, when and if he wants to acquire liquidity later on, he will be more confident in investing his money in the first place. Because high-frequency traders are always looking to enter into a trade, even when the potential return on investment is minimal, they effectively function as market makers in contemporary financial markets. They therefore also provide most of the liquidity. The empirical data bear these claims out. In a 2011 study, Hendershott, Jones & Menkveld found that algorithmic trading improves liquidity and enhances the informativeness of quotes. While their study focuses on algorithmic trading rather than HFT specifically, and while their method of investigation is somewhat indirect (they use changes in market structure that increases AT as an exogenous instrument to measure the causal effect of AT on liquidity), their findings are generally borne out by subsequent studies. (Sun, Kruse & Yu, 2012) These studies have shown not only that under certain optimization models HFT can supply liquidity, but also that it can supply liquidity while reducing overall execution cost. Therefore, there seems to some evidence both for HFT increasing liqudity in financial markets, and for its ability to lower overall transaction and execution costs.
- Improved pricing efficiency
HFT increases pricing efficiency by reducing bid-ask spreads. A bid-ask spread is the amount by which the ask price exceeds the bid price for an asset in the market. It is essentially the difference between the highest price that a buyer is willing to pay for an asset and the lowest price that a seller is willing to accept. Bid-ask spreads exist naturally in financial markets due to informational asymmetries between buyers and sellers. Such spreads are thought to be a negative thing because they reduce pricing efficiency in the market.
HFT is thought to reduce this bid-ask spread by exploiting these differences at such a high-frequency the market quickly and automatically corrects the discrepancy where, without HFT practices, these discrepancies would have taken longer to correct. Again, the data seem to bear this hypothesis out. Hagströmer and Norden (2013), in a study of the Stockholm financial index, found that bid-ask spreads were tighter with the presence of market-making high-frequency traders than without them. Similarly, Brogaard and Garriott (2018) found that when 11 HFT firms were introduced into a new Canadian stock exchange called Alpha over four years, the bid-ask spreads converged to those on the Toronto stock exchange as more and more HFTs were introduced. In other words, the bid-ask spreads decreased as the pricing mechanisms became more efficient as more HFT’s were introduced. The data do suggest there is some positive correlation between the presence of HFTs and the pricing efficiency of a financial market.
Disadvantages of HFT
Despite these benefits of HFT, there are potential disadvantages as well.
- Higher systematic risk
HFT is thought to increase systematic risk in a market due to the sheer volume of trades engaged in. High-frequency trades account for a substantial proportion of dollar trading volume and liquid capital. This creates greater systematic risk in a market because market stability becomes more and more dependant on the stability of these trades. If something goes wrong in HFT trading mechanisms, it has the potential to cause shockwaves throughout financial markets. This phenomenon was seen clearly enough in the Flash Crash of 2010, for which HFT was generally held responsible. But do the data support such an attribution of blame?
Various studies do indicate that HFT has a non-trivial effect on market volatility. In a 2016 study, Virgillio found that the participation of high-frequency traders leads to a statistically significant increase in volatility when the market was under stress. On the other hand, when the market was not under stress, no abnormal behaviour resulting from HFT was found. Similarly, Dover (2019) says that HFT plays a role in market melt-downs or melt-ups, which is exacerbated when the markets have low levels of liquidity. So HFT does play a role in increasing market volatility, and this role is especially prominent when the market is already under stress or is facing low-liquidity or high-volatility scenarios.
- Disadvantages to other investors
HFT tends to hurt investors who do not use algorithmic or high-frequency trading methods. Because of the substantial profits HFT traders tend to make, especially as an aggregate of thousands of trades with very small margins, HFT methods tend to ‘crowd out’ traditional investors who do not have access to the same kind of information and sophisticated technology. Studying the S&P 500, researchers found that high-frequency traders “made an average profit of of $1.92 for every contract traded with large institutional investors and an average of $3.49 when they traded with retail investors. This allowed the most aggressive high-speed trader to make an average daily profit of $45,267,” according to the 2010 data. (Parker, 2013) The researchers concluded that these profits came at the expense of other, traditional investors, and that such practices may lead these investors to leave financial markets and abandon traditional investment techniques.
The fact that HFT crowds out traditional investors may not seem like a disadvantage to some. After all, if it increases market efficiency and reduces transaction costs, why does it matter whether or not it harms traditional investment practices? Shouldn’t these inefficient practices be done away with anyway? The problem with such an argument is it ignores the systematic inequality such an outcome creates. Because HFT methods are only accessible to those investors with the relevant technological expertise — often large financial institutions — the proliferation of HFT leads to the accumulation of most of the liquidity, asset capital, and investment portfolios in the hands of a few. The death of the traditional, independent investor will be the death of any semblance of equitability in the financial market. It will lead to increased inequality and concentration of financial power. And neither of these are a good.
- Market manipulations
Because HFT allows enormous profits to be made on tiny price variations, there is some worry that its proliferation will lead (or has led) to greater market manipulation. Investors may be able to make huge profits on small, hard-to-detect manipulation, like the kind caused by floating rumors. One study found there was a positive correlation between the use of HFT and ticking – movements in the price or price quotation of a security – and a negative correlation between the use of HFT and Price Dislocation Alerts – an alert indicating asset mispricing. (Frine & Lipone, 2012) These two metrics typically proxy for market manipulation, and positive correlation with HFT for both may have lent some credibility to the claim that HFT leads to increased market manipulation. However, the negative correlation between the use of HFT and Price Dislocation Alerts suggests that HFT actually leads to less asset mispricing, not more. Either that, or the mispricing that HFT exploits is so small and so quick that it does not register an alert. Nevertheless, there does seem to be room for HFT practices to benefit from market manipulation, even if it hasn’t occured – or hasn’t seemed to occur – on a wide scale just yet. In order to prevent potential market manipulation, regulators need to find a way to construe and apply current financial regulations regarding market manipulation to the practice of HFT. In order to do that, they need devise a method for providing proof of such manipulation within the current legal framework. (Niwa, 2016)
Conclusion
What do these empirical findings say about the future of HFT practices? Are they a positive sign, a cause for concern, or an area for further inquiry? It seems that the advantages of HFT outweigh the potential disadvantages. While market volatility is a definite risk, the relevant government regulation can ensure that HFT’s capacity to increase volatility during times of stress be reduced. This has been done effectively enough by MiFIR, but there is room for improvement that should be pursued. A disadvantage of HFT that is a cause for serious concern is the effect of this practice on smaller, more traditional investors. It is not clear how such an outcome can be prevented by government regulation. The death of the traditional investor seems to be an inevitable outcome of the proliferation of algorithmic trading practices in general and HFT in particular. Still, the negative consequences of this outcome can be mitigated by ensuring that HFT does not create anti-competitive monopoly power in financial markets. This can be done by placing a cap on the amount (measured in dollar trading volume) of high-frequency trades that financial institutions engage in. By mitigating the disadvantages of HFT, we can make the greatest use of its tangible advantages.
Works Cited
Brogaard, Jonathan & Garriott Corey. “High-Frequency Trading Competition.” Journal of Financial and Quantitative Analysis (2018). Available online at: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2435999. Accessed June 22nd, 2019.
Cheng, Evelyn. Just 10% of trading is regular stock picking, JPMorgan estimates. Available online at: https://www.cnbc.com/2017/06/13/death-of-the-human-investor-just-10-percent-of–trading-is-regular-stock-picking-jpmorgan-estimates.html. Accessed on July 2nd, 2019.
Dover, Stephen. “Volatile Markets: Are High-Frequency Traders to Blame?” Available online at: www.advisorperspectives.com/commentaries/2019/02/27/. Accessed on June 23rd, 2019.
Niwa, Daisuke. Market Manipulation Using High Frequency Trading and Issues Facing Japan. (2016) Japan Lawyers Guide. Available online at: https://media2.mofo.com/documents/161000-high-frequency-trading-japan.pdf. Accessed on July 4th, 2019.
FINRA, the Financial Industry Regulatory Authority. Regulatory Notice 15-09: Equity Trading Initiatives: Supervision and Control Practices for Algorithmic Trading Strategies (March, 2015). Available online at: www.finra.org/sites/default/files/notice_doc_file_ref/Notice_Regulatory_15-09.pdf Accessed on June 20th, 2019.
––––– FINRA Rule 3110 (Supervision) (April 2014). Available online at: http://www.finra.org/industry/supervision. Accessed on July 2nd, 2019.
Frino, Alex & Lipone, Andrew. The impact of high frequency trading on martket integrity: an empirical examination. (May, 2012) Commissioned by the UK Governemnt’s Foresight Project. Available online at: https://pdfs.semanticscholar.org/4644/0e5fc9339f52c5f1f857b1ffd7e8073e2bf8.pdf. Acccessed on July 3rd, 2019.
Gomber, Peter. “High-Frequency Trading.” SSRN Electronic Journal (January, 2011). Available online at: http://ssrn.com/abstract=1858626. Accessed on June 20th, 2019.
Hagströmer, Björn, & Norden, Lars L. “The Diversity of High Frequency Traders.” SSRN Electronic Journal. Available online at: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2153272. Accessed: June 19th, 2019.
Kenton, William. Flash Crash. Available online at: https://www.investopedia.com/terms/f/flashcrash.asp. Accessed: July 1st, 2019.
Menkveld, Albert J. “The Economics of High- Frequency Trading: Taking Stock.” Annual Review of Financial Economics 8 (2016): 1-24.
Norton Rose Fulbright. MiFID II / MiFIR: High Frequency and Algorithmic Trading Obligations (2014). Online publication. Available at: www.nortonrosefulbright.com/en/knowledge/publications/6d7b8497/. Accessed on June 23rd, 2019.
Parker, Tim. “Has High-Frequency Trading Ruined the Stock Market for the Rest of Us?” (January, 2013). Available online at: https://www.investopedia.com/financial-edge/0113/. Accessed on June 23rd, 2019.
Sun, Edward W., Kruse, Timm, & Yu, Min-Teh. “High Frequency Trading, Liquidity, and Price Impact.” SSRN Electronic Journal (January 2012).
Virgillio, Gianluca. “The Impact of High-Frequency Trading on Market Volatility.” The Journal of Trading (Spring, 2016).
Weller, Benedict & Bruno, Michelangelo. “Is EU regulation of high frequency trading stringent enough?” LSE Business Review. Available online at: https://blogs.lse.ac.uk/businessreview/2018/10/08. Accessed on June 20th, 2019.
Zhang, Frank. High-Frequency Trading, Stock Volatility, and Price Discovery (2010). Available online: http://ssrn.com/abstract=1691679. Accessed on June 19th 2019.
Image Courtesy of Ten Minute Millionaire Insider