Automated trading and its effect on the markets

Goldman Sachs introduces a computer programme to allow investors to trade bonds without speaking to any member at the investment bank. It could increase the frequency of efficient trading activities and reduce the transaction costs as it does not depend on the Bloomberg platform any more. Moreover, there are other advantages of using mathematic algorithm to make trading decisions, as all trading orders executed by computers will allow the certain rules that are pre-programmed. Therefore, the irrational decisions or miscalculation will not happen in such trading process and human errors could be limited to almost zero. In addition, the trading executed by machines has been proved to be more profitable than human activities, as based on researches on the top traders, the machine has a better performance than the human traders on average. Therefore, computing will replace human forces in the field of trading including highly frequent trading.

If computing has been widely used in the markets, then we could probably see that trading could concentrate at certain periods as I do not think there could be very different ideologies and strategies behind the trading algorithms, instead of creating a very different strategy, investment banks are more likely to improve their calculation efficiency in order to make faster decisions and place orders ahead other investors. Based on such assumption, the trading activities could become concentrated. The weakness of such system is it is vulnerable to unexpected risks in the markets. Moreover, the standard of cybersecurity could be pushed up to a much higher standard. In addition, once a mistake is made, the losses could be extremely high and almost impossible to be recovered. Furthermore, as the vast majority of trading is executed by machines and the market is still a zero-sum game, worse algorithms will be eliminated because of their performances very fast and if the speed of developing new algorithms is not fast enough to replace the eliminated algorithms, then the algorithms used by the market could be more and more similar and could potentially increase the system risk, as the risk is not dispersed.

To conclude, the use of machines in trading could increase the profits of the investors who use better algorithms; however, such system may be more vulnerable to unexpected risk, as our human traders may make some experienced decisions which could potentially limit the losses when facing such risk.