Reflections on “Artificial Intelligence In Finance”
The article “Artificial intelligence in finance” has allowed me to develop a big-picture understanding of what artificial intelligence and machine learning can do for trading. This newfound understanding has inspired me to study both quantitative trading and options trading strategies, as well as how to leverage machine learning tools such as classification, SVM( Support Vector Machine ), and regression to enhance my trading performance.
I found the explanation on the application of statistical and machine learning methods for finance the most significant within the paper. Machine Learning methods are more about algorithms rather than asymptotic statistical processes; they are great for processing nonlinearity, but not linear data. Machine Learning emphasizes high dimensional prediction problems but not statistical inference, which is better for solving low dimensional problems (Buchanan, 2019, p. 19).
The applications of unsupervised learning and supervised learning for the financial markets also gave one more insight on investing in general. The unsupervised Machine Learning clusters algorithms and develops topic models: clustering algorithms can be used to understand traders’ behavior, while topic models can help us understand the behavioral drivers of market participants. For example, we can use text mining and Natural Language Processing to analyze news, which allows us to track the positive and negative changes of certain stocks’ economic direction. Consequently, the investor can react quickly, or the Artificial Intelligence can place the order in response to the change (Buchanan, 2019, p. 20).
Another important category was supervised machine learning models. Supervised Machine Learning encompasses predictive analytics, random forests, neural networks, and support vector machines. The random forests model is useful for forecasting future business performance. Neural networks can assess corporate financial health, and are especially convenient because they require fewer assumptions, but can achieve a higher degree of prediction accuracy than most statistical methods. Neural networks can be used in sample pricing and delta hedging, as well as in the Black-Scholes formula. Support vector machines are very useful for making long/short equity trading decisions (Buchanan, 2019, p. 23).
Applications of artificial intelligence and machine learning within finance are still in their early stages, and have the potential to play a more significant role in the financial markets in the future. Artificial Intelligence techniques still need to be tested under different special situations such as financial crises, as Artificial Intelligence methods still have high risk in the financial market due to the uncertainty factor. I believe there will be a boom in Artificial Intelligence trading and similar services in the next few years and hope to be an active participant in the movement.
Written by Frank Chen & Edited by Vivian Fang