Business, Investment planning

AI in Financial Services: the Future of Stock Forecasting?

22 June 2023

The modern world has provided us with a host of innovative ideas, designs and technology that has the potential to make the next 100 years vastly different from the last. The wave of smart technology – especially phones, tablets and laptops – has created a digital world that we are all, to a degree, plugged into. As ground-breaking as this was, the human race does not stand still, and it seems the Artificial Intelligence wave is here to stay.

What is Artificial Intelligence (AI)?

The Oxford Dictionary defines ‘artificial intelligence’ as ‘the theory and development of computer systems able to perform tasks normally requiring human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages.’ Platforms such as ‘ChatGPT’ and Google’s ‘Bard’ have provided casual users with the ability to get opinions, solve maths questions and even write whole university essays with referencing. The AI boom seems as though it is here to stay and, despite its critics, it feels like it’s just the beginning.

Artificial Intelligence (AI) in Financial Services

An outline image of a connected brain to showcase the power of AI.

So, how does AI affect us in the financial service world? Well, considering ChatGPT has shown it can display the ability to understand headlines from financial news and how they might impact stock prices, there is a feeling that AI may even be able to predict stock movements. As ChatGPT has advanced language capabilities programmed into it, the software can detect nuances and subtleties within headlines and stories, allowing it to make informed stock market predictions.

Alejandro Lopez-Lira – a professor at the University of Florida – and his assistant, Yueha Tan, recently published a paper explaining how Large Language Models (LLMs) can be useful in predicting stock market prices1. Using ChatGPT to examine news headlines through ‘sentimental analysis’, Lopez-Lira and Tan found that the chatbot was able to predict the direction of the next day’s returns by determining if the headlines are ‘good or bad’ for a stock.

Fed over 50,000 headlines about public shares listed on the Nasdaq, New York Stock Exchange and the American Stock Exchange into the chatbot, ChatGPT assessed the stocks returns during the next trading. At the time of the research, the bot could only process data from September 2021 onwards, so the sample period was from October 2021 until December 2022. Recent updates to ChatGPT now mean that premium users can access data from the live web, potentially enhancing the AI’s ability to forecast stock prices.

According to the results of the experiment, Lopez-Lira and Tan concluded ‘Our results suggest that incorporating advanced language models into the investment decision-making process can yield more accurate predictions and enhance the performance of quantitative trading strategies’.

How AI Could Influence the Financial Industry

Lopez-Lira and Tan’s experiment had many conclusions, but there are two major findings that could be influential for AI in the finance world moving forward:

  • The results of the study can potentially benefit asset managers and institutional investors by providing evidence of the ability of LLMs to predict stock market returns. Moving forward, they can use this as a basis to incorporate AI into their work or not.
  • The research can help regulators and policymakers understand both the benefits and risks of increasing the presence of LLMs in the financial market. This can help create regulatory frameworks and fair governance of AI in finance.
Graphic of three light blue cogs to represent the possible connectivity between AI and the financial industry.

Lopz-Lira also predicted that the more AI is integrated into stock market predicting, the more efficient the market will become. With this in mind, the return predictability of ChatGPT could likely be very low in 5 years.

Limitations of AI for Stock Forecasting

A financial chart in an uptrend with a dollar sign covering the middle section, coloured in brown.

There are plenty of limitations to the professor’s experiment. For one, ChatGPT never actually analysed target prices or did any calculations – the experiment analysed headlines and language, with Lopez-Lira saying that ChatGPT has less than a 1% chance of a correct prediction without a headline. This goes to show that the technology is not perfect and, in the case of stock forecasting, goes to show all results are contextual.

Worse yet, LLMs have a recent history of struggling to crunch complex mathematical questions. Google’s ‘Bard’ famously answered a maths question incorrectly at their keynote address, whilst CNET’s AI-written articles recently came under criticism for making basic mistakes that an adviser would know, but an AI bot would likely skim over. These may seem like small and ineffectual in isolation but continued overlooking of minor details in this industry can affect people’s money and, in turn, their livelihoods.

Exploring Further Impacts of AI on the Financial Sector

As well as the influence ChatGPT and other LLMs may have on stock forecasting in the future, it is also important to consider the other impacts AI will have on the financial services industry. Beyond stock forecasting, AI’s role in the financial sector extends to other areas such as customer relationship management. Although AI won’t be used exclusively in the financial sector to improve customer support, financial organisations face a greater challenge compared to other industries due to the highly sensitive nature of personal information involved when managing customer data. This demands the highest data security measures to ensure that AI not only enhances customer experience by creating a more personalised service but also protects personal financial information from security risks or data breaches.

Ethical Considerations of AI in the Financial Industry

When it comes to dealing with personal data, one of the main concerns about the use of AI in the financial sector relates to ethical issues. Financial institutions must ensure that AI systems are transparent, and fair, whilst not discriminating against certain groups of customers. It is a contentious issue amongst artificial intelligence use broadly, but AI models must be carefully trained and closely monitored to avoid bias or discrimination.

Understanding Consumer Preference for Financial Advice

It’s also worth asking what consumers really want: financial advice from a real human who can understand the complexity of people’s concerns or dialogue from a generated bot that’s been trained to regurgitate financial advice with no financial stake of its own. The exciting new possibilities of AI are exactly that – exciting and possibilities – but not a new reality. Yes, these are tools that can enhance the efficiency and output of financial services as a whole, but there is no replacement for the grounded advice of a financial adviser, identifying and empathising with you as a person. Speak to one of our expert financial advisers today.

The levels and bases of taxation, and reliefs from taxation, can change at any time and are generally dependent on individual circumstances.

The value of an investment with St. James’s Place will be directly linked to the performance of the funds you select, and the value can therefore go down as well as up. You may get back less than you invested.

1 Can ChatGPT Forecast Stock Price Movements – A. Lopez-Lira (April 2023)

Sign up to our mailing list to get the latest wealth and mortgage and protection insights

Mailing list

What news would you like to hear about?(Required)