Are Machine Learning And AI the Future of Investing?

Nowadays, we are hearing machine learning and artificial intelligence almost everywhere, I remember watching my partner playing with an astrology application that supposed to tell one’s fortune this month, in 6 months or even the whole year according to the "big data" and "artificial intelligence" embedded behind it. In recent years. these technologies have developed so fast and been broadly marketed across all different industries. The proliferation of the concept has people already react: “what’s the big fuss?" when talking about AI. But do we really understand it and how it applies? I'd say NO with a high level of certainty. That's why I feel passionate to study them in both academic and empirical manner and share my humble opinions on how these technologies have affected and transformed the investment industry, where I work as an insider, over the past few decades. 

 

Brief History of Machine Learning and Artificial Intelligence in Financial Industry

The buzzword "Artificial intelligence", officially received recognition in 1956 after World War 2, was created at the Dartmouth Summer Research Project on Artificial Intelligence (DSRPAI) hosted by John McCarthy and Marvin Minsky. 

In 1960s, the Bayesian models, which was used heavily in machine learning, was applied in auditing areas to offer decision maker objective and rational probability to make more accurate assessments in decision making. 

During 70s to early 1980, the development of AI has seen slowdown due to the deduction of government funding. However, in the financial services industry, James Simons’ quantitative hedge-fund Renaissance Technologies made its debut and currently managing about $165 billion in April 2021.

Then boom till now. AI has swept over the financial industry and numerous models were created afterwards in order to reduce cost of operations, predict market patterns, and assist better investment decisions, etc. 

 

Why AI have grown so fast in Financial Industry

Efficiency and Speed

When it comes to financial forecasting and company fundamental analysis, AI could be very efficient to facilitate the process, which is usually cumbersome and time consuming. Trust me, I started as a fundamental analyst, and no one wants to spend hours to hard collect and process data from different sources, and double check the data used was correct. Moreover, people in the Wall Street have this crazy desire for speed of everything, especially the speed of processing information. It usually means that the faster process one can achieve on a new piece of information, the more likely he or she could be the first one who takes a large piece of bread.

Cost Reduction

It is not only embodied on the low transaction cost we enjoy today, but it enormously lowers the service cost if someone works with a financial advisor. I remember the time when Wall Street broker charge 1% on each transaction they execute for their clients (imagine one buys a million-dollar worth of stock, the broker makes $10,000 right away by move a mouse and click.) Money is so easy to make, right? Nowadays this business model will go out of business if the company does not innovate, while almost all competitors offer much lower or sometimes $0 transaction costs. This is very important and has make investment very affordable for individual investor.

Phase out Cognitive and Emotional Biases

In general, doesn’t matter if someone is a hedge fund manager or an individual investor, they are facing the same problem everyday: am I making an optimal and rational decision? Does my investment decision possess any cognitive errors or emotional biases?

·       Cognitive errors (information is not processed correctly) are handled relatively easily given the right education and instruction of problem solving.

·       Emotional biases are inherited from people's temperaments and are much harder to deal with.

By building an AI based model, managers can moderate the negative impact of emotional biases in the investment decision making, such as day to day buy/sell securities orders. In fact, many managers are applying a hybrid approach of combining human-judged decision facilitated with machines.

 

How AI would Affect Personal Investor

New Era with More Options

As far as I can think, for individua investors, one of the biggest disadvantages is that they do not have much time following the markets and companies like financial professionals do. AI and Robo-advisors have become very popular among retail investors, especially younger generations who do not want to spend much time on personal investing and would like to reach certain financial goals. We are living a time when personal investors have more options to choose instead of working with a financial advisor who might charge a hefty fee for managing their funds. 

 

Obstacles we have seen in the application of Machine Learning /AI 

Data Quality & Cleaning Process

I have asked for insights with friends who are great data scientists working on quantitative trading systems and modeling, the feedbacks I usually get are:

·       Modeling results often contradict with the historical results while they are testing their algorithmic system. It turned out that they need another 3-4 data scientists to clean out the data and study it if the data is fit to use in the first place. Otherwise, it could turn out to be a "garbage in, garbage out" situation.

·       It is relatively easy to train the model and produce alpha (the return that is better than the return of S&P500 index) with in-sample (selected historical prepared data) data. However, when it comes to out of sample data (all available and new data) that the results are more important and decisive, the accuracy of modeling predictive capability have gone down, or sometimes the results are meaningless.

The Pursuit of Alpha

From an empirical point of view, we have agreed that the machine can bring somewhat stable return relative to index, and benefits of diversification in a different period, but it is hard to produce outsized return (we call that Alpha) as it is usually achieved by human judgement and execution. Because of the complexity of data in financial markets and the amount of new data generated every day, there has not been an all-weather system to “smartly” beat the market. Will there be one soon? I don’t know, but I am always optimistic about the future.

 

The application of machine leaning, and AI have indeed transformed the financial/investment business for the last few decades. I’d like to make a bold speculation that the intermediary (brokers, traditional traders) of investment business would be largely wiped out and replaced by a more cost-efficient and intelligent machine or system (maintained by human still) in this coming decade. It is cruel for some professionals who have done their business in a more traditional way, but it is part of history of evolution. Financial safety and capability are an essential part of everybody's life, I believe that it is always wise to learn from a macro perspective and act in micro way to achieve financial goal. 

 

 

 

References:

https://analyzingalpha.com/history-of-ai-in-finance