top of page

Using AI to Analyse Company Earnings Calls

Updated: Jan 29


Company earnings calls are one of the most important sources of information for investors and analysts. These calls, usually hosted quarterly, provide an opportunity for management to discuss the company's financial results, provide updates on strategic initiatives, and field questions from analysts. However, these calls can often be long and filled with technical jargon, making them hard to decipher for many individuals. That's where artificial intelligence (AI) comes into play. With its ability to process large amounts of data quickly and accurately, AI can provide valuable insights from these calls. In this article, we'll delve into how AI is used to analyze company earnings calls.



The Role of AI in Processing Earnings Calls


Transcription and Text Analysis: The first step in the process is transcribing the call. AI-powered speech recognition technology can convert spoken words into text with remarkable accuracy. This text can then be analyzed using natural language processing (NLP), a subfield of AI that focuses on the interaction between computers and human language. NLP allows AI to understand, interpret, and generate human language in a valuable way.


Topic Modeling: Another application of NLP is topic modeling, which can identify the main topics discussed during the call. This can be incredibly useful for investors and analysts who may be interested in specific aspects of the business, such as product development, market expansion, or financial performance. Topic modeling algorithms, such as Latent Dirichlet Allocation (LDA), identify patterns of word use and group them into topics. This can help analysts quickly understand the main points of discussion without having to listen to the entire call.


Predictive Analytics: AI can also be used to make predictions about future performance based on the information discussed during the earnings call. By using machine learning algorithms, AI can identify patterns in the data that might not be obvious to humans. This can include correlations between specific words or phrases and subsequent stock price movements.


Sentiment Analysis: Sentiment analysis, also known as opinion mining, is a natural language processing technique used to determine the sentiment or the emotional tone behind words. This technique is extensively employed in AI analysis of earnings calls to gauge the underlying sentiment of the statements made by the company's management and analysts. Traditionally, analyzing the sentiment in an earnings call relied on human analysts who would listen to the call and interpret the tone and mood of the speakers. This process was not only time-consuming but also highly subjective, with different analysts potentially arriving at different conclusions. With the advent of AI, it's now possible to automate this process, providing a more objective and quantitative measure of sentiment. AI algorithms can analyze the words used in an earnings call, determining whether they convey a positive, negative, or neutral sentiment. Advanced algorithms can even detect more complex emotions like excitement, frustration, or disappointment. In addition to analyzing the words themselves, AI can also take into account the context in which they're used. This is particularly important in earnings calls where the same word might convey different sentiments depending on the context. For example, the word "challenge" might be negative when used in the context of financial results, but positive when referring to market opportunities. Sentiment analysis can provide valuable insights into the company's performance and the market's perception of it. For instance, a positive sentiment from the management might indicate optimism about future performance, while negative sentiment could suggest potential issues or challenges. However, while sentiment analysis is a powerful tool, it's not without its challenges. AI algorithms can struggle with subtleties of human language like sarcasm or irony, and they lack the human ability to understand the broader context outside the call. Despite these limitations, the use of AI for sentiment analysis in earnings calls represents a significant advancement over traditional methods, providing a faster, more objective, and more data-driven approach to understanding the complex narratives in earnings calls.


Benefits of Using AI to Analyse Earnings Calls


There are several benefits to using AI to analyze earnings calls. These include:

  • Speed and Efficiency: AI can process and analyze an earnings call in a fraction of the time it would take a human. This allows for quicker decision-making and a more timely response to new information.

  • Quantification of Qualitative Data: Earnings calls often contain a wealth of qualitative information that can be hard to quantify. AI can convert this qualitative data into quantitative data that can be easily analyzed.

  • Bias Reduction: AI can help reduce bias in analysis by treating all data objectively. While human analysts might be influenced by their perceptions of the company or its management, AI simply analyzes the data as it is.

  • Detailed Insights: AI can provide more detailed insights than a human analyst might be able to. For example, it can identify subtle changes in sentiment or topic emphasis that might be overlooked by a human.


A Case Study


AI technology has proven highly effective in analyzing company earnings calls, providing insights that were previously difficult to quantify or even unnoticed. One case study highlighting the use of AI in analyzing earnings calls is the Japanese investment bank, Nomura. Nomura used AI technology to analyze earnings call transcripts and found that the sentiment expressed by analysts during the calls was a more reliable predictor of a stock's performance than the management's remarks. According to Joseph Mezrich, head of quantitative strategy at Nomura, companies with the worst analyst sentiment in earnings calls significantly underperformed the market in the days leading up to the call and even more dramatically underperformed following the call. Conversely, those with positive analyst sentiment performed better​. Nomura used a technique called "natural language processing" for its analysis, utilizing an AI platform known as ProntoNLP. They created two portfolios of stocks: one based on analysts' sentiment from the earnings calls and the other based on management's sentiment. The portfolio measuring analyst sentiment outperformed the management sentiment portfolio by about 25 percentage points, further highlighting the accuracy and relevance of analyst sentiment​​.


Challenges and Considerations


While AI offers numerous benefits, there are also challenges and considerations to keep in mind. One of the main challenges is ensuring the accuracy of the AI's analysis. While AI can process large amounts of data quickly, it's only as good as the data it's trained on. If the training data is biased or incomplete, the AI's analysis may be as well. Additionally, while AI can analyze what is said during an earnings call, it may struggle to understand what is not said. For example, it may not understand the implications of a pause or a change in the speaker's tone. It's also important to note that AI lacks the contextual understanding that human analysts have. For instance, AI might struggle to understand industry-specific jargon or subtle sarcasm. Moreover, there is also the risk of crowding in the market if the same AI analysis methods are used widely. As these services become popular, they may become less effective as investors flock to the same trades. For example, if all investors act on the same insights from an AI analysis of an earnings call, the advantage that these insights provide may diminish over time​.


AI offers a promising solution to the challenge of analyzing company earnings calls. By leveraging technologies like natural language processing and machine learning, AI can quickly and accurately process these calls, providing valuable insights into the company's performance and the management's sentiment. However, it's important to be aware of the limitations of AI and to continue to improve its accuracy and sophistication in response to market dynamics. With the right approach, AI can be a powerful tool for investors and analysts, helping them make more informed decisions based on a comprehensive and objective analysis of company earnings calls.


 

The interest in artificial intelligence is intensifying in corporate America, with references to AI and related terms during calls with investors more than doubling from a year ago. Companies are eager to show they're at the forefront of ways to use algorithms that can parse vast amounts of information based on simple prompts. Despite the growing alarm from industry leaders and governments at the potential risks of the new technology, companies haven't been deterred from voicing their commitment.

39 views0 comments

コメント


bottom of page