Crash Landing on Wall Street

Investigating the Causal Relationship Between Crashes and Market Performance for Boeing 737!

Authors : Devina Chhajer, Diya Patel, Ishika Gupta, Kaitlyn Ho, Tanvi Lamba

Description: Boeing is the leading manufacturer of commercial aircrafts and supplies to majority of the airline companies today. If you are travelling by air, there's a high chance you are travelling in a Boeing aircraft. So, for such a company to have frequent crash incidents, emergency landings and part blowouts is quite concerning. With this project, we are investigating the economic and social impact these kind of incidents have by performing Causal Analysis, Sentiment Analysis and Topic Modelling techniques on the data we collected for 3 specific incidents. A special focus is laid on analyzing Crash 2 by performing causal, stock price stickiness, topic modelling and sentiment analysis for company and government releases following this event.

Incidents covered:

1. Crash 1: Lion Air Flight 610 on Oct 29, 2018 - no survivors
2. Crash 2: Ethiopian Airlines Flight 302 on Mar 10, 2019 - no survivors
3. Incident 3: Alaska Airlines Flight 1282 on Jan 5, 2024 - door blowout

Datasets: The merged_dataset that can be found in the 'Data' folder is an excel workbook consisting of data scraped from news and social media articles, Stock data, financial data and orders and cancellations data. We scraped the data using the extension scraper.io.

Data Sources:

1. Textual data: News, Quora, Reddit, FAA statements, Boeing Website.
2. Financial Data: Yahoo Finance, Boeing Website, Airbus Website
3. Operational Data(Orders and Cancellations): Boeing website

Code Files: There are 2 code files and 1 html file in the 'Code' folder.

1. File "Final_Code (latest)" is a python notebook consisting of all the analysis done so far on the dataset. It is best run on Google Colab.
2. File "Ethiopian_Airlines_Vader_Score" is a R file consisting of the code for calculating the vader scores for sentiment analysis. An html file with the same name can also be found.

Interesting Insights:

1. As expected, there was a ripple effect of incident 1 and 2 on incident 3 as pre-incident, public sentiments were less positive and more negative for each following incident.
2. Price fall stickiness was for 11 days after crash 2.
3. All 3 incidents happened on a weekend.
4. In 2020, 90% of the total Boeing cancellations were for the 737 models.
5. Discussions towards supporting those affected by the Lion Air incident suggested that proactive efforts to address customer concerns and demonstrate commitment can yield positive outcomes.
6. Despite the gravity of crash 2, both public sentiment and stock prices began to regain strength about 1 week later due to the above mentioned point.

Blog: We have posted a blog on Medium, giving an overview of the study and explaining our interpretations for each analysis. Following is the link to the blog. Link --> https://medium.com/@devinachhajer24/crash-landing-on-wall-street-9e8d11cc8bc5