The recent proliferation of trading algorithms has elevated the potential efficiency of financial markets. Much to the dismay of market participants, it has elevated potential volatility even more.
On Monday, the 5th February 2018, the US stock market saw its first sharp correction in years, with share prices falling more than 10 per cent. The days of low volatility in markets are over and a far less predictable environment is likely to be the norm. It is highly controversial to assess whether or not computer driven algorithms triggered the recent stock sell-off; however, it is undisputed that algorithms, derived from statistical models, are increasingly replacing human beings. Although alpha-creating algorithms create significant benefits by providing sufficient liquidity, they can also have an adverse impact on the marketplace as a whole. Ought we fear losing human control in the market or can we assume that economic formulas presupposing human rationality are for once valid?
Andy Haldane, the chief economist of the Bank of England, once pointed out: “A US trading firm became insolvent in 16 seconds when an employee inadvertently turned on an algorithm. It took the company 47 minutes to realise it had gone bust”.
Algorithms have become so ingrained in our financial system that markets could not operate without them anymore. Their influence ranges from the replacement of the middlemen in a transaction at the most basic level to the sophisticated analysis of stock performances, financial statements and newsfeeds. At their best, algorithms could create an efficient capital allocation machine, predicting prices and movements rationally. Algorithms could even lead to more stability in the market as they provide sufficiently high liquidity and theoretically push prices back to reasonable levels as soon as valuations become unsustainable. Psychological, social and emotional impacts in the decision-making process would hence be wiped out. It is our irrationality that was and still is responsible for numerous crises and sell-offs as well as bubbles, with the crypto-hype being only the most recent example.
At its worst, however, interconnected and interacting algorithms can easily spiral out of control. They are barely regulated, hardly predictable and can accentuate declines by selling a large number of stocks automatically once certain price levels are reached, hence intensifying the movement even further. The implementation of the Markets in Financial Instruments Directive (MIFID) is undoubtedly a cornerstone of EU law and deliberately tries to regulate algorithmic trading in order to prevent market distortions. However, in order to ward off potentially dangerous market disruptions from happening, it will be necessary to initiate further legislation. Besides that, traders cannot monitor the algorithm while they execute trades, reinforcing a lack of transparency.
“What´s most concerning is that an event in one market ripples through to global markets, raising volatility to historic levels”, explains R. Brown, author of Chasing the same signals.
Incorrect algorithms pose a fundamental risk to the entire financial system. In the case a faulty algorithm which is trading billions, the resulting chain reaction from a simple fault could be immense.
Algorithms play a fundamentally important role in financial markets. It would be utterly misguided to call algorithmic trading as a whole into question. The most prominent development facilitating unprecedented growth and ultimately prosperity over the last centuries is derived from powerful computing, high connectivity speeds and technological disruption. Hence, it would be inappropriate to replace them. What it needs however, is more regulation within that aspect and a change of the skill-set of traders. Fundamental valuation allied with daily news will incrementally be replaced by back-testing and number-crunching. Mr. Market could soon come to an end. It is difficult to assess, however, whether or not Mr. Robot will finally stabilise markets and lead to a market based on rationality.