At Bewgle we apply our NLP capabilities on any unstructured data, using our patented AI models, to generate actionable insights from it. Bewgle’s Natural language processing, machine learning models, fundamentally analyze the text and output the answers to questions, the insights, topics, sentiment, adjectives and other key features that we promise to our customers.
Here we discuss Bewgle’s capability to analyze financial sentiments using company earnings calls over a period of time.
An earnings call is a conference call between a public company’s management team and interested stakeholders (investors, analysts, etc.) to discuss earnings/financial summary for a specific period. Analysts/investors use information from the earnings call in their fundamental analysis of the company. Earnings calls are considered one of the key resources for investors and equity analysts.
Analysing Apple Inc’s earnings call transcripts for Q3 & Q2 for 2022 and Q3 & Q2 for 2021
Input source – https://wallmine.com/nasdaq/aapl/transcripts
The Bewgle platform analysed Apple’s performance for Q2 and Q3 2022:
- Summarized earnings calls for a quarter
- Compared earnings calls across quarters
- Provided responses to custom questions
Summary for quarter
Bewgle analyzes the data and extracts topics and their sentiments. For every topic detected by the system, here is what is identified:
- Snippets from the text that give positive sentiments are displayed as ‘Endorsements’.
- Snippets with negative sentiments are displayed as ‘To Improve’.
Using these snippets from the top influential topics, Bewgle projects the summary for the specific quarter, including positives and negatives from that quarter.
Summary for Q2 2022 produced by the system:
Summary for Q3 2022 produced by the system:
As seen above, the summary captures the essence of the quarter. It highlights these aspects:
- Record revenue figures
- Products in demand
- Year-over-year change
When compared with the actual transcripts of earnings calls, these summary points are highly accurate w.r.t the sentiments.
Comparison across quarters
As stated above, after analyzing the data, Bewgle extracts topics and their sentiments from the text. For every topic, the system calculates the sentiment score and ranks it based on the frequency of usage (number of times it appears in the text) and the sentiment. Using these top influential topics, data from different time periods can be compared with each other.
Comparing earnings calls across quarters:
Q2 2022 vs Q3 2022
Quarter, Customer, Growth, Revenue, Service, iPhone, Base, and Apple are the top topics detected by the system. The dots indicate their sentiments for the respective quarter.
Conclusion: Q2 was better than Q3.
The Bewgle summary seen in the previous section also had the same conclusion.
This Summary and Comparison are produced out of the box, using the topic sentiments and insights produced by the system.
On the Bewgle platform, the ‘AnyQ’ feature gives users the opportunity to ask different questions about the data to get more insights. It then points the users to the exact text that answers the question.
Here are the questions posed to the system along with Bewgle’s responses:
1. Which products saw strong demand?
Q3 2022 – In 2022 Q3, iPad, wearables and iPhones were the products in high demand.
Q3 2021 – In 2021 Q3, iPad, Mac and wearables were the top-performing products.
‘Wearables’ was in the second position in Q3 2021, but went to the third position in 2022.
This was produced in the highlights of Q3 2022 as well – “Wearables, Home, and Accessories revenue was $8.1 billion, down 8% year over year”
2. What is the impact of Covid?
Based on Bewgle’s responses for 2021 and 2022 Q3, we learn that Apple Inc is coming out of Covid impact.
Q3 2021 –
Q3 2022 –
3. What is the impact of the Russian war?
The Russian war response is for the time period of Q2 2022.
The AnyQ feature gives us the capability to not only gain topic insights and sentiments but to also ask custom questions to get additional insights.
Bewgle’s analysis achieved these three goals:
- Summarized earnings calls for a quarter
- Compared them across different quarters
- Received answers from the system for custom questions
This use case can be extrapolated for any company’s earnings calls and financial statements and for analyzing/comparing financial sentiments over long periods of time.
Gaining such an insightful understanding of the unstructured data of earnings calls will be hugely useful for analysts and investors who can analyze multiple sources of financial data over various periods with efficiency.
If you have a similar use case or would like to try Bewgle on your unstructured data, please reach out to us at email@example.com .