Algorithmic Trading: Scapegoat or Legitimate Issue?

Trading algorithms and computer-driven systematic trading have continued to be the scapegoat for many along Wall Street, both in the face of the coronavirus and otherwise. These fancy new forms of technology are earmarked as the biggest risk to the market at large, with Michael Lewis citing high-frequency trading as the “rigging” of the stock market, and figures as prominent as Secretary of the Treasury Steve Mnuchin believing that they have a negative impact on market turmoil (Cox, 2018). However, is this caricature accurate? While even those that program trading algorithms themselves recognize the potential downsides of implementing this technology, the broad use of computers as the only scapegoat is misguided and deceiving (Wigglesworth, 9 Jan 2019).

It is important to make a distinction between the different types of quantitative trading. The most infamous is the previously mentioned high-frequency trading, which relies on blindingly fast financial data to exploit minute changes in prices and momentum, or to leverage their speedier latency to inflate prices and sell stocks to buyers in between the milliseconds of the trader pressing send and the transaction occurring (Kindig, 2020). Other forms of “systematic trading” are less focused on such granular data and exploitative practices, but instead are programmed to identify and adjust to specific trends within the market. Whether it be traditional technical patterns or a complex mathematical model that can trade based upon fundamental analysis, trading algorithms have continued to thrive in such a dynamic and diverse market, and they have continued to evolve in how they attack the market.

The distinctions between different types of trading algorithms is very important, as some of the criticisms of quantitative trading are very valid, but they only apply to specific types of trading. High-frequency trading, for example, is designed specifically to capitalize on price discrepancies in a fashion that has genuinely destabilized the markets (Scholtus, 2014). HFT funds are at their most active during periods of high short-term volatility wherein they rapidly buy and sell risky investments with considerable negative impacts for the rest of the market in turbulent times (Wigglesworth, 9 Jan 2019; Wigglesworth 30 Sep 2015). This strategy definitely deserves our criticism, and high-frequency trading should definitely not be grouped together with other forms of systematic investing.

Although the trades executed by algorithms are obviously done without human insight, outside of high-frequency trading, algorithms are designed to think the way that humans think, and they capitalize on very legitimate strategies. One of the most popular forms of systematic trading is called “risk parity”, which is essentially a strategy that chases a specific level of volatility across a diverse portfolio of assets. So, as volatility goes up (an indication of the market entering a downturn), risk parity algorithms will trim their exposure to different assets, potentially exacerbating the downturn as a whole, in theory (Banerji, 2020). There are still others that trade on the momentum in the commodity markets, and others that are programmed to identify deviations from the fundamental value of different assets. Thus, many of the driving theses in systematic trading are based in trading strategies that human beings have capitalized on for years. Selling off riskier assets to reduce exposure to a volatile market did not begin with computers, but rather has its roots in the human tendency to avoid risk. Furthermore, algorithmically following trends and fundamental mispricing are the building blocks of technical and fundamental analysis. Having these strategies executed by computers does not make them less legitimate than when they are executed by humans.

The evidence of a direct correlation between spikes in volatility and algorithmic trading is not strong, as many investors are griping about rebalancing and trend-following without actually considering the tangible effects (Groth, 2011). But, that is not to say that algorithmic trading has not come to fruition during a period of tumultuous markets. Since the financial crisis, markets have seen events like the Flash Crash and other swift, temporary corrections across markets. Now, the Flash Crash has directly and correctly been linked to high-frequency trading, but in terms of other sudden spikes of volatility and prices, the direct connection to computers making the decisions, rather than humans, is empirically false (Groth, 2011). Rather, the sudden adjustments in volatility exposure or quick changes in which trends to follow is a reflection of the explosion in access to information in the Internet age. As computing power has grown more robust, it is simply easier and more competitive to have computers scan for the exact same strategies that you yourself would hold. Instead of basing your technical decisions on hourly or daily data, computers can easily consume minute data to take advantage of trends. Thus, the prominence of algorithmic trading, as well as volatility, is moreso a sign of the times, rather than a direct cause and effect relationship.

Furthermore, an objective positive that has come with systematic trading has been the increased liquidity in the market. Many of the algorithms that currently navigate the market monitor liquidity on a multitude of securities, and they have grown to be prime influences in the supply and demand of liquidity itself (Hendershott, 2013). Ultimately, algorithms tend to consume liquidity when it is cheap and create liquidity when it is expensive (Hendershott, 2013). They have a strong presence in narrowing the spreads for many large stocks, helping the market at large to see the effective market price by demanding the necessary liquidity to find the efficient price at any point in time, at a much quicker rate than regular human trading (Hendershott, 2011). Thus, in theory, enhancing the ebbs and flows of the market could contribute to volatility, in a flash-crash situation, but it also reflects all of the information on market conditions across the board.

In conclusion, are algorithms truly exacerbating the problem of liquidity? Not really, for a couple of different reasons. First of all, making all of these trades instantly reflects the nature of computers in the modern day. There is a demand for quick decisions like nothing traders have ever seen, and algorithms are the best way to achieve this. Furthermore, these are algorithms that ultimately mimic the human mind. They are not AI robots that are completely detached from humanity, but instead are trained to act exactly like humans. Thus, a potential solution could be regulation that prevents specific investment strategies, like exclusively following momentum and trends or high-frequency trading. However, eliminating computer-executed algorithms is not the right choice.

Works Cited:
Banerji, Gunjan and Zuckerman, Gregory. “Why Are Markets So Volatile? It’s Not Just the Coronavirus.” The Wall Street Journal, 16 Mar 2020. https://www.wsj.com/articles/why-are-markets-so-volatile-its-not-just-the-coronavirus-11584393165

Cox, Jeff. “Mnuchin: Computerized trading ‘definitely’ helped drive this week’s big market swings”. CNBC, 6 Feb 2018. https://www.cnbc.com/2018/02/06/mnuchin-computerized-trading-definitely-had-market-impact.html

Groth, Sven S., "Does Algorithmic Trading Increase Volatility? Empirical Evidence from the Fully-Electronic Trading Platform Xetra" (2011). Wirtschaftsinformatik Proceedings 2011. 112.

Hendershott, Terrence; Jones, Charles and Menkveld, Albert. “Does Algorithmic Trading Improve Liquidity?” The Journal of Finance, Vol 66, No. 1. Feb 2011.

Hendershott, Terrence and Riordan, Ryan. “Algorithmic Trading and the Market for Liquidity.” Journal of Financial and Quantitative Analysis, Vol 48, No.4, pp. 1001-1024 Aug 2013.

Kindig, Beth. “New Age of Stock Market Volatility Driven by Machines.” Forbes, 10 Apr 2020. https://www.forbes.com/sites/bethkindig/2020/04/10/new-age-of-stock-market-volatility-driven-by-machines/#196567026dda

Lewis, Michael. Flash Boys. Penguin Publishing, 2015.

Scholtus, Martin; van Dijk, Dick, and Frijns, Bart. “Speed, algorithmic trading, and market quality around macroeconomic news announcements.” Journal of Banking & Finance, Volume 38, Pages 89-105, Jan 2014. https://doi.org/10.1016/j.jbankfin.2013.09.016

Wigglesworth, Robin. “Volatility: How ‘Algos’ Changed the Rhythm of the Market.” The Financial Times, 9 Jan 2019. https://www.ft.com/content/fdc1c064-1142-11e9-a581-4ff78404524e

Wigglesworth, Robin. “Markets: Can They Really Be Tamed?” The Financial Times, 30 Sep 2015. https://www.ft.com/content/aad452a8-660b-11e5-a57f-21b88f7d973f