New Study Suggests AI Can Outperform Analysts in Predicting Company Results

Artificial intelligence has already made a significant impact on our lives in the last decade and it is anticipated that it can play a huge role in the financial sector as well. A recent study has shown that the Large Language Model (LLM) can even outperform financial analysts in detecting a company’s future prospects.

The researchers used a unique process where the model helps to break a complex task into simple smaller bits and then analyze them. By using this process, LLMs achieved a high accuracy rate of more than 60% while detecting future earnings. This is a full 7 percentage points higher than the average analyst prediction, according to the study. Researchers only fed these AI LLM models with balance sheets and income statements and based on that the model was able to derive the prospect. This shows how powerful AI capabilities have become in today’s world.

AI vs Human Analysts:

AI has never been on par with human analysts in the past however with this new innovative process called chain of thought prompting, AI now has the power to even outclass the human analysts with higher accuracy rates. The study also found that the LLMs’ forecasts added more value when human biases or inefficiencies – such as disagreements among analysts – were present. There are also scenarios where AI was not able to perform effectively. This is while analysing small firms. This is likely because context tends to matter more when making predictions for smaller or more variable firms.

The study found that while both LLMs and human analysts had trouble with these more complex financial circumstances, human analysts were generally better at dealing with them. This is likely because they can factor in soft information and context found outside of the financial statements.

Generative AI vs Predictive AI: Check Key Differences Between them

The results of the study indicate that AI will probably have a bigger role in making financial decisions in the future. According to the authors, “GPT and human analysts work well together, rather than replacing each other.” This means that the future of finance will likely involve humans and AI working together to make the best decisions.

Some researchers even wrote, “Taken together, our results suggest that GPT can outperform human analysts by performing financial statement analysis even without any specific narrative contexts,”.

Thus, the AI models used in the study were better at predicting future earnings than some financial analysts, which means they could be a useful tool for investors. However, the study also suggests that the future of finance will involve humans and AI working together to make the best decisions in the complex financial markets.

What are the key differences between large language models (LLMs) and generative AI?