Unlocking value from text data
Natural Language Processing (NLP) is a sub-field of Data Science to process and analyze unstructured text data.
In 2019, Barclays was already using machine learning tools to analyze company managements’ commentary in earnings calls. As the technology continues to gain momentum, we can expect to see more companies benefiting from NLP in the future.
Do the highest-paid CEOs speak the same way in Earnings Calls?
In this article, we will try to identify and analyze word patterns of the three highest-paid CEOs (Tim Cook, Elon Musk, Tom Rutledge) through their commentary in their respective earnings calls.
The dataset used in this article consists of Earnings Calls in Quarter 4, FY2020 of Apple, Charter Communications, and Tesla. Filtering was done on the dataset so that only commentary by the CEOs are used in the analysis.
Product vs. overall growth-focused conversations
First, let’s look at the most commonly used words by the CEOs.
Based on the bar chart below, Apple and Tesla’s CEOs had more similar word patterns. Cook and Musk focused on sharing the more important/latest features of the company’s product. On the other hand, Charter Communications CEO Rutledge mentioned no specific product or service and focused on the growth side of the company.
These findings could suggest that Apple and Tesla have a bestseller plan to bring in more revenue. We could also explain that Charter Communication seeks to retain its current business plan rather than bring in any game-changing plan.
Uncovering competitive advantages/shortcomings through sentiments
Another interesting aspect of Natural Language Processing is to categorize words into sentiment categories. The charts below label the CEOs' commentary into sentiments that are important in the financial context.
Do remember that corporate executives may try to sound more upbeat in earnings calls. Hence, we should analyze earning calls with due diligence.
The three CEOs mentioned similar wordings regarding requirements, challenges, and dependency on other matters that all companies face.
Using CEO’s sentiments to determine if the company is a growth or value stock
If we take a close look at the wordings in the positive and negative sentiment categories, we could make a few assumptions on what the CEO thinks about the company performance.
Cook and Musk both mentioned the word “excited” in their earnings call meeting and Rutledge used the word “optimistic”. Using these two sentiments, we could make a high-level presumption that Apple and Tesla are in rapidly growing industries (potentially growth stocks) and Charter Communication is in the mature industry.
Word sentiments and competitive advantages
Cook also talked about creativity and innovation, whereas Musk mentioned efficiency and improvement. The mentioned themes could mean that the company has a plan in the segments or plainly what the company is not doing well with.
The wordings could mean that Apple and Tesla are trying to bring added value to their customers through different strategies. In the future, we may see Apple employing/acquiring intellectual and knowledge assets and Tesla achieving greater profit margins by improving the company’s operational efficiency producing call batteries and Model 3.
Machine learning says three themes stood out in the CEOs commentary
Besides using lexicons and frequency to analyze text data, we could also apply machine learning techniques to uncover hidden insights from text.
The chart below shows the cluster of words used by the CEOs that the LDA algorithm has grouped. If we take the first word, with the highest % in each group, we could presume that the themes involved in the commentary are: video, customer, and growth.
For the “video” topic, Tesla and Charter Communications use video technology in their product, Tesla uses video labeling in its autopilot system whereas Charter Communications sell video subscriptions to its customers.
In the second topic, “customer” was on top of the list. This list also has a mixture of words used by Musk (“car”, “drive”, “cell”) and Cook (“apple, “service”, “store”). We can assume that this list is related to Apple and Tesla’s customer service.
As for the last topic with the word “growth” on top of the list, we can make the hypothesis that Apple’s business growth has a significant dependency on how well the company sells its iPhone products. Growth for the three companies may also depend on their products and how well the CEOs bring the company to a worldwide level.
- CEO speech patter in earnings call may be either product or overall company growth-oriented
- It is common that CEOs share in the earnings calls that the company success depends on other organizations in its eco-system
- We could assume based on the CEO’s sentiment if the company is in a growing industry by comparing words such as “excited” and “optimistic”
- The three themes that emerged from unsupervised machine learning when applied to Apple, Tesla, and Charter Communications CEOs speech pattern was video, customer, and growth