Skip to content
Home » Machine Learning and Money Markets

Machine Learning and Money Markets

ChatGPT was recently known to help predict market movement, but Artificial Intelligence has been doing that job for much longer.

  • Since the development of modern-day computing, AIs have been at the heart of Wall Street.
  • One mistake in 2018 caused AIs to cause a meltdown in global markets.

Although Artificial Intelligence (AI) has become a recurring theme, it had a role in financial markets for years. They have even caused some disasters along the way. Recently, a news item was making the rounds on the internet about the possibility of using ChatGPT to predict market movements. The reality? Thanks to its capabilities, it is possible to use this AI to create predictive models. Many people said that the advent of AIs would completely change the markets. Although, other AIs already do the job and are much better.  

AIs are not of one type; they depend on their application and the way they learn. ChatGPT, for example, uses a Deep Learning system. But unlike many other AIs, it has an easier way of communicating with people thanks to the GPT (generative pre-trained transformer) engine, which can use natural human language.  

In addition to Deep Learning, there are AIs with Natural Language Processing (NLP), which can understand and interpret human communication. There are also Machine Learning AIs, which can identify patterns within a conglomerate of data.

Intelligent algorithmic trading

With the growth of computing in the early 2000s, the capabilities of AIs increased. They went from being able to beat world chess champion Garri Kasparov to predicting market moves very accurately. People name these tools “algorithmic trading” as they are automated systems that create and close thousands of orders per second on stock exchanges worldwide. 

According to BBC technology reporter Padraig Belton, AI algorithms processed three-quarters of all New York Stock Exchange in 2018. Automated AI trades drive the majority of the market. The advantage that an AI has over a human trader is processing power. NLP-type AIs can process thousands of news stories worldwide in a matter of minutes, quickly finding a possible cause that can define the trend of a stock.  

AI to predict market movements

Large financial firms already use AI to manage their funds. There is a reliance, almost absolute, on what an AI will recommend. JP Morgan is one of the leading examples of this. In 2016 it became known that the company would start the development of COiN, an AI that allows analyzing large volumes of data by applying a natural processing language algorithm. The AI directs to both the financial and legal fields.  

Some data revealed that JP Morgan’s AI could perform the equivalent of more than 360,000 person-hours of finance work in a few seconds. In 2018, the CEO of this company stated for CNBC at the Delivery Alpha conference that COiN had helped process more than 20,000 contracts on its platform.  

Another example is Blackrock, the largest financial conglomerate on the planet, with investments in multiple items, including bitcoin (BTC). This company has a subsidiary dedicated to market trading research using pure AI, which has developed its electronic system known as Aladdin. This system uses AI with Deep Learning and Natural Language Processing to measure and calculate market trends in advance. Data reveals that in 2021, Aladdin predicted market movements with profits close to $1 billion.

AI and disasters in financial markets

While AIs seem a panacea in financial markets, all could be better. AIs know no human emotions; they operate based on data and patterns, which can lead to some problems. Between 2010 and 2012, there was a phenomenon that experts called a flash crash. These stock market collapses endure for minutes or seconds but profoundly affect a financial need because they liquidate traders’ positions. What are the causes of these flash crashes? AIs.

Since an AI only acts automatically, once it detects a trend, it will start trading in favor of that trend. In a flash crash, an error causes the AIs to start trading downwards, causing the rest of the AIs to replicate this behavior and causing a “crash” in the market. Upon detecting the error, the AIs themselves correct the trend, although it is usually too late, even if it is a correction of seconds. The latter is because this movement can liquidate billions in positions since it is unexpected and goes against market logic.  

The origin of the error that can initiate a flash crash varies, but it can usually be false information. AIs base their learning on verifiable sources, but what happens with corrupted data sources? This situation occurred in 2013 when hackers took control of the Associated Press (AP) Twitter account. It tweeted about an alleged attack on the White House, bringing down the Dow Jones index by more than 130 points. All in a matter of seconds.  

Usually, these AIs are under human supervision. However, specific regulations try to prevent AIs from taking complete control of the markets, causing global pandemonium. One such regulation is the U.S. Securities and Exchange Commission (SEC) Rule 15c3-5, which requires broker-dealers to implement risk controls to limit certain types of trading.   

ChatGPT can pave the way for investment.

Understanding the point of AIs in the markets, ChatGPT’s arrival in these scenarios may mean something other than a revolution regarding market predictions and so on, as other, much more experienced AIs do such work. However, this new AI does open up an opportunity, and it is education-based. Financial markets are complicated and unpredictable. However, ChatGPT’s raison d’être facilitates understanding by explaining complex topics in simple language, bringing financial markets closer to ordinary people, and promoting education within this context.

Leave a Reply