Artificial Intelligence and the Financial Economy
Artificial intelligence (AI) is flourishing. Its growth is driven by a combination of technology and market factors: 1/ The computing power of phones and laptops is much greater than that of many central computers used half a century ago. Investors have always used several types of technology to glean market information ahead of their competitors. The speed of information and communication technologies has allowed those who can afford it to have an advantage over their competitors. 2/ The data available for analysis are growing. 3/ Smarter and more reliable algorithms have been developed in recent decades. 4/ The use of this data has been miraculous in solving health problems, which has given new impetus to the use of AI in other sectors. But, as with any innovation, there are institutional blockages. For example, there is a lack of skilled labour available in this area and there are concerns about the inadequacy of data privacy standards.
Economic benefits and fears have made AI a political and economic issue at the international level. Given its great market potential, the major bloc of countries (the United States, China, Canada and the European Union) are vying for primacy and, at the same time, accentuating its growth.
Financial markets are now managed by computer: human intervention is diminishing; traders are being replaced by transactions executed electronically. Not only are these faster, but they are instantly saved and added to the information database. The use of high-frequency trading allows for profiting from the slightest anomalies. The next to disappear are portfolio managers. Securities analysts are being replaced by algorithms and passive investment funds. Today, the passive funds tracked by Morning Star are worth more than those managed by humans (The Economist, 2019).
At the end of the last century, we already saw the emergence of quantitative hedge funds that analyze the market using data. But only a quarter of these funds are managed solely by human managers. Much more important are those who use artificial intelligence. Some quantum funds (hedge funds using quantitative investment techniques), such as Bridgewater and many of BlackRock’s funds, use algorithms to perform data analysis, but leave humans to select transactions. This is because AI-driven algorithmic investment often identifies factors that humans do not; but human analysts may seek to understand what the machine has spotted to find new “explanatory” factors. But many quantum funds, such as Two Sigma and Renaissance Technologies, as well as some BlackRock funds, push automation even further, using machine learning and artificial intelligence (AI) to enable machines to choose the shares to buy and sell. The speed with which they make decisions and execute them gives these investors an advantage.
The result is that the stock market is now extremely efficient, but volatile. New markets serviced by robots are significantly reducing costs. The lower cost of executing a transaction means that new information about a company is instantly reflected in its price. Major consumer brokerage sites expect trading costs to be zero. These lower fees contribute to the liquidity of securities. More liquidity means a smaller spread between the price a trader can buy a stock and the price he can sell it.
According to a recent study (Research and Markets, 2019), the global AI chip market was worth $6.6 billion in 2018. This is expected to reach $91.2 billion by 2025. Currently, North America leads the AI chip market. Chips are used in autonomous vehicles, healthcare, cybersecurity, display screens and smart clothing. The main player in North America is the United States, the leading player in Europe is the United Kingdom and the leader in the Asia-Pacific region is China.
The risks are many. The first fear is that major technology players (such as GAFAM in the United States – Google, Amazon, Facebook, Apple, Microsoft – or BATX in China – Baidu, Alibaba, Tencent and Xiaomi) will dominate the market and oust historical operators such as banks as well as small-FinTech startups. Such an increase in monopoly power means that customers will ultimately have less choice and pay more. From a more technical point of view, the risk is related to piracy associated with the concentration of a large mass of data in the hands of a small number of actors.
Burgundy School of Business, Dijon, France