The age of automated trading is upon us. Most of the trading today is happening to, from and between computers that have very clever programs to consistently and reliably invest in the stock market better than any human could. High-frequency trading, also referred to as HFT, gradually developed since 1983 after NASDAQ introduced a purely electronic form of trading. Since then, an increasing number of firms dived into this new market opportunity where revenues are defined by millions of small trades per day with a few pennies as the profit for each transaction. In this highly competitive financial niche, competition is measured in Internet speed, computing power, and incredibly smart algorithms.
High-frequency traders are using an experimental type of network cabling and infrastructure to speed up their systems by billionths of operations per second, the latest move in a technological arms race to execute stock trades as quickly as possible. These new optic fiber cables are already installed in the basements of the NYSE (New York Stock Exchange) to allow HFT firms for almost real-time, delay-free transactions. Indeed, the old continent and Asia are already connected to the United States via trans-oceanic high-speed Internet cables and companies in the AI-based trading game are springing up like mushrooms worldwide.
What about the algorithms behind this outstanding technology? An interesting social phenomenon that is shaking the American job industry is the impressive salary of quants. Quants are, according to new estimates, the highest-paid workers in the US ($500 000/yr), even before specialized surgeons. This relatively new position’s job is to develop and implement complex and effective models used by firms to make financial and business decisions about issues such as investments and pricing. These models are then implemented in what represents the competitive advantage for every HFT company, its algorithms. Such algorithmic programs are let crawling free on the Internet, and they do great at analyzing everything related to finance, analyzing platforms ranging from the WallStreetBets subreddit to the Bloomberg and Financial Times pages.
Other than applying sophisticated strategies such as price arbitrage and statistical arbitrage, artificial intelligence also conducts something called sentiment analysis: thorough a series of keywords found in financial-related content on the web, computers decide whether the price of a specific security will rise or fall and act accordingly. One of the most impressive examples of what massive computing power can achieve is the Hewlett-Packard case. On the 18th of August 2011, Bloomberg posted an article related to an important multibillion-dollar deal regarding HP. The American multinational, just 4 seconds after the article was out on the Internet, saw a positive 13.9% rise in its stocks. Akin speed of execution is beyond human capabilities of processing information. The only limit for these bots is the speed of data through fiber optic cables, thin air and computers’ components.
What happened in August 2011 was the second big news covering the effects of high-frequency stock trading. Just one year before, on May 6, 2010, the American stock market ran into what was later called the crash of 2:45 or simply the flash crash. In a matter of minutes, the Dow Jones Industrial Average (an index that traces the 30 biggest American companies) faced almost a 1000 points loss: more than a trillion dollars in capitalization were erased from the market. Authorities determined that high-frequency traders sold aggressively to eliminate their positions and withdrew from the markets in the face of uncertainty. From a 2014 investigation, it emerged that HFT firms did not cause the Flash Crash but contributed to it by demanding immediacy ahead of other market participants. It is also interesting to highlight what some Wall Street analysts suggest. They calculated that HFT, despite its controversies, helped absorb the shock and did not either cause or contributed to the negative trend. According to them, retail investors panicked more than usual, and the market would have fallen way faster if it wasn’t for algorithms that decelerated what seemed like an unstoppable free fall. Provided that determining the effect of one single player in the trading game is no easy job, financial experts all agree that HFT has one major positive impact on the bourses. Whensoever a retail investor wants to buy or sell a security, there is always liquidity to sustain that transaction; in the majority of cases, when traders in flesh and bones pay or get paid for exchanging securities, they can be sure that it is not a he or a she that is buying or selling but rather an “it”, a bot, an algorithm for some HFT firm that conducts countless operations a day.
Regardless of the premises that make HFT firms look like some shady business or illegal practice, they are nevertheless real companies and as such they incur in costs, very high costs. Technology is exceedingly expensive, and the environment is, to say the least, competitive, with firms battling each other for that billionth of a second of faster computing operation. This is not to mention the struggles that high-frequency trading businesses face in Europe in these months, marked in particular by a strong increase in energy prices. In defiance of what just said, the biggest fear of all for these firms is none other than the government. All around the world, authorities are fighting HTF practices by taxing single trades more and more (in Italy there is a 26% flat tax on trades while in the US the highest percentage traders can pay is 23.8%). Needless to say, companies incur in something called micro-taxes (taxes that on their own seem insignificant) due to the millions of different exchanges taking place 24/7. Anyway, as we all know, the law struggles to keep up the pace with technological progress, therefore making it a little bit easier for high-frequency traders to evade some taxes.
The high-frequency trading phenomenon is getting more attention than ever and it is important not to underestimate it. Recently, a firm called Hudson River Trading was described as responsible for around 5% of the New York Stock Exchange trading volume, which is more than one billion shares exchanged every single day. These impressive results are just the beginning of a century in which financial markets, just as every other aspect of our lives, will be shaped by complex AI. With the upcoming 5G technology, retail traders will be once and for all forced to abandon their already risky profession. In the end, the only humans to succeed in the financial arena will be long-term investors, those who remain impassible before the bourses’ ups and downs and keep their little fortunes for decades to come.