TL;DR:
- The University of Auckland researchers have utilized artificial intelligence (AI) to evaluate the actual value of companies based on various factors.
- Their machine learning algorithms outperform traditional methods in stock valuation accuracy.
- Undervalued stocks identified by AI tend to rise in price, providing investors with profitable opportunities.
- The researchers also developed AI algorithms to identify peer firms in cases where traditional methods fall short, which is particularly useful in countries like New Zealand.
- Determining the value of a company can be influenced by human bias, making the AI methodology game-changing.
- Industry professionals often value companies by comparing them to others in the same industry, but this process can be subjective and biased.
- The researchers used tree-based machine learning models to allocate firms based on fundamentals, identifying close peers with similar characteristics.
- Their models analyzed a vast sample of US common equities, resulting in more accurate valuations that resemble the true value of a company.
- Dr. Helen Lu and Dr. Paul Geertsema from the University of Auckland spearhead this research, contributing to the field of AI in finance.
Main AI News:
In a groundbreaking study published in the prestigious Journal of Accounting Research, esteemed academics Dr. Helen Lu and Dr. Paul Geertsema from the University of Auckland’s Business School unveil how artificial intelligence (AI) is reshaping the assessment of companies’ true worth. By employing machine learning algorithms, they demonstrate that these cutting-edge models provide more precise stock valuations than conventional methods.
Lu and Geertsema’s pioneering machine learning technique surpasses traditional models in terms of accuracy, effectively identifying undervalued stocks that subsequently experience price appreciation. This presents investors with enticing opportunities for profit without assuming additional risks.
Moreover, the researchers have developed machine learning algorithms that assist professionals in identifying peer firms when conventional methods prove inadequate. This innovation is particularly invaluable in countries such as New Zealand, where the identification of relevant peers can pose challenges. Lu, an expert in utilizing AI to solve financial problems, highlights the significance of this advancement.
Assessing a company’s value involves subjective decision-making, exposing stock prices to the influence of human bias. According to Lu, this machine learning methodology is a game-changer due to its ability to counteract such biases. Professionals in the industry typically evaluate companies by comparing them to others within the same sector. However, this process is prone to biases, and evidence suggests that practitioners strategically select peers to achieve desired valuation outcomes.
“For instance, when valuing a software company, industry professionals typically seek peer firms in the technology industry that offer similar products. However, this process remains highly subjective,” explains Lu. “Defining the ‘tech industry’ precisely and determining comparability between companies is challenging. Factors like pricing power and growth potential should also be considered. Skilled professionals often rely on hunches influenced by personal biases.”
To mitigate these issues, the researchers developed and trained tree-based machine-learning models renowned for their efficacy. These models automatically determine the optimal allocation of firms to different leaves in a tree based on fundamental characteristics. Consequently, businesses frequently assigned to the same leaf are considered close peers with similar fundamentals. Conversely, firms rarely allocated to the same leaf possess different underlying traits.
The researchers’ models extensively analyzed a vast sample of common equities listed on major U.S. stock exchanges, including NYSE, NASDAQ, and AMEX, spanning from January 1980 to December 2019. The final dataset utilized by the machine learning models consisted of 1,811,785 firm-month observations, encompassing 16,201 firms. This framework has the potential for application in stock markets worldwide.
Lu emphasizes that their approach not only generated more accurate valuations compared to traditional models across different firms and over time but also yielded valuations that more closely aligned with a company’s true value.
Dr. Helen Lu serves as the FinTech leads for the Master of Business Analytics program, while Dr. Paul Geertsema teaches Financial Machine Learning and Data Analytics for the MBA curriculum, both at the University of Auckland. Their pioneering research showcases the transformative potential of AI in revolutionizing the field of stock valuation, opening new doors for investors and industry professionals alike.
Conlcusion:
The groundbreaking research conducted by the University of Auckland’s esteemed academics, Dr. Helen Lu and Dr. Paul Geertsema, signifies a significant leap forward for the market. By harnessing the power of artificial intelligence and machine learning algorithms, the evaluation of companies’ true value becomes more accurate and less prone to human bias. This advancement not only enhances stock valuation accuracy but also presents investors with lucrative opportunities by identifying undervalued stocks that have the potential for price appreciation.
Moreover, the development of AI algorithms to identify peer firms addresses a critical challenge faced by industry professionals, particularly in regions like New Zealand. Overall, this transformative research paves the way for a more sophisticated and informed market, empowering investors and industry experts to make well-informed decisions based on more accurate valuations.
