Can AI forecast market trends and outsmart portfolio managers?


Artificial intelligence (AI) is currently making waves throughout society and portfolio management is no exception. Where is AI already having a practical impact? I asked ChatGPT how it sees AI affecting the work of portfolio managers.

AI can help portfolio managers with a key problem faced in the industry: When do you have sufficient data needed for decision-making in order to minimise the uncertainty in making those decisions?

AI can already be used to analyse and summarise large amounts of text, such as prospectuses for the issuances of new securities, and to collect and analyse publicly available documents from different sources. It can also be used to assess investor sentiment, for example by analysing data from social media. At its best, AI can also be used to discover causal relationships that affect markets and thus help develop algorithms and trading strategies. And, of course, it speeds up all these processes described above.

While AI can already be given decision-making power, it still needs to be guided. AI applications are not yet ready to just plug-and-play, so to speak, but require a user's vision of how AI will be properly utilised and continuously developed. At this time, AI is still more of a ‘virtual colleague’ that challenges and spars with its own actions, and which the user has to develop at the same time. Even self-learning systems need to be guided according to one’s own objectives.

How ChatGPT predicts the future of portfolio management

While writing this post, I became interested in what ChatGPT thinks about the interplay between portfolio management and AI. In the following, there is our discussion in full.

TH: The success of portfolio managers has been compared to throwing darts. Can AI do better?

ChatGPT: The work of portfolio managers is demanding and requires constant analysis and decision-making. Artificial intelligence can provide portfolio managers with significant help in many tasks, such as data analysis and rapid decision-making. AI can efficiently search and analyse large amounts of data, but the final decision may require human intuition and experience. Portfolio managers can benefit from AI to assist them, but human skills and judgement are still keys to success.

TH: Can an investment robot / AI application speculate?

ChatGPT: AI applications can analyse markets and make forecasts, but speculation is often based on high-risk, short-term trading. Most AI applications focus on long-term investing and risk management rather than speculation. Speculation involves a high degree of risk, whereas AI is better suited to help investors make calculated and well-informed decisions.

TH: Do you believe that in the future AI applications will compete with each other in the market?

ChatGPT: Yes, the market for AI applications is growing, and competition between them is likely to intensify. Companies are constantly developing new and improved AI solutions for investors. This offers investors more options, but at the same time emphasises the importance of the right choice and use of AI.

TH: There is an old saying for investors which goes, “the trend is your friend”. Can AI spot trends before anyone else?

ChatGPT: AI can analyse historical market data and detect trends and anomalies. It can also use complex algorithms to identify potential trend reversals. Although AI can be faster and more efficient at detecting trends than humans, it may not always be completely error-free in forecasting.

TH: Can AI correctly assess the valuation levels and expected returns on investments?

ChatGPT: AI can perform complex analyses and estimates of valuation levels and expected returns on investments. It is based on data and mathematical models, but it is important to note that all estimates are based on historical data and forecasts, which can change rapidly. The final decision requires human judgement and a careful consideration of the risks involved.

TH: When investment bubbles arise, you can really make money. But can AI still work in an irrational market?

ChatGPT: AI can help identify potential bubbles or market volatility, but irrational markets are often difficult to forecast. AI can provide analytical insights, but investors still need to be cautious and consider the risks, particularly during bubbles.

TH: The market can sometimes be wrong, but it is difficult for an individual investor to go against the market. Can AI beat the ‘invisible hand’ that guides the market, as made famous by Adam Smith?

ChatGPT: AI can enable investors to make better decisions and identify potential inconsistencies in the market. It can also help investors manage risk more effectively. However, markets are complex and susceptible to many factors, and perfect profits are not guaranteed. AI can improve the possibilities for decision makers, but it does not necessarily "beat" the market, but rather help investors make better decisions within it.