Quantile Mechanics (Aviation Stress - Market Volatility/M&A)

Inspiration

Air travel generates a large volume of operational data, but very little of it is used in financial analysis. While exploring aviation datasets, we noticed that many airlines operate on the same routes. On busy routes, several carriers compete for the same passengers, gates, and airport capacity. This kind of overlap often leads to congestion, delays, and thinner profit margins for everyone involved.

This made us wonder whether that operational pressure could actually be measured and used as a signal. In particular, we wanted to explore two ideas: whether route competition between airlines could hint at potential mergers or acquisitions, and whether unusual disruptions in aviation — such as spikes in delays or cancellations — might ripple into financial markets. Since flight delay data reflects real-world disruptions like weather events, demand shocks, or supply chain issues, we hypothesised that unexpected aviation stress might help explain short-term market volatility.

What it does

Our project converts raw U.S. aviation data into two signals that may be useful for understanding market behaviour and airline industry dynamics. The first signal analyses route overlap between airlines to identify potential merger pressure. By comparing the routes operated by each airline, we measure how similar their networks are and identify carriers that compete heavily on the same city pairs. High overlap suggests stronger competitive pressure, which can make consolidation or mergers more likely. The second signal focuses on aviation system stress and market volatility. We build a stress index using metrics such as flight delays, cancellations, and severe delay rates across major airport hubs. After removing normal seasonal patterns, we isolate the unexpected portion of the stress signal and test whether spikes tend to precede movements in financial markets like crude oil futures or equity indices.

How we built it

We began by building a data pipeline to collect and process aviation data from the U.S. Bureau of Transportation Statistics. These datasets include detailed information about routes, delays, and cancellations over multiple years. After cleaning and standardising the data, we aggregated operational metrics from major hubs to create a daily aviation stress index and separated the expected seasonal patterns from abnormal disruptions. For the merger analysis, we constructed route networks for each airline and compared them using similarity measures such as the Jaccard score to quantify how much their routes overlap. We then built a backtesting engine to test whether spikes in aviation stress could predict financial market movements across different assets and time lags. The results were visualised through an interactive dashboard. Our tech stack included Python, with libraries such as Pandas and NumPy for data processing, Matplotlib and Seaborn for visualisation, Streamlit for the dashboard interface, and Jupyter notebooks for exploratory analysis and testing.

Challenges we ran into

One of the main challenges was cleaning and standardising the raw aviation data. The BTS datasets span several years and are distributed in different formats, with inconsistent column names and occasional missing records. Significant preprocessing was required before we could build a reliable dataset for analysis. Another challenge was accounting for seasonality. Aviation activity changes throughout the year, with predictable spikes during holidays and summer travel. Without adjusting for this, the signal would mostly reflect the calendar rather than meaningful disruptions. We also had to carefully align aviation data with financial market data to avoid timing errors or look-ahead bias in our backtests.

Accomplishments that we're proud of

One of our biggest achievements was turning an unconventional dataset into a meaningful analytical signal. Aviation delay data is widely available but rarely used in financial modelling, and our project shows that it may contain insights about both market behaviour and industry structure. We are also proud of building a complete end-to-end pipeline: from downloading raw government datasets to producing backtested strategies and an interactive dashboard. The system allows users to explore aviation stress trends, analyse airline route overlap, and experiment with strategy parameters in a reproducible way.

What we learned

This project showed us how powerful alternative data sources can be. Many publicly available datasets contain useful signals that are overlooked simply because they fall outside traditional financial data sources. We also learned that unexpected deviations often matter more than raw levels. Aviation activity naturally fluctuates with seasons and holidays, so isolating the abnormal portion of the stress signal was crucial. Once we removed the seasonal component, the remaining signal provided a clearer view of disruptions that may influence markets.

What's next for Aviation Stress Market Volatility and Merger Acquisition

One of the next steps would be to automate the pipeline so the aviation stress index updates automatically as new flight data becomes available. This would allow the system to generate real-time alerts when unusual disruptions occur. We also plan to expand the analysis to international flight data and explore additional alternative datasets, such as real-time flight tracking or airport traffic statistics. Combining multiple data sources could strengthen the signal and help us better understand how disruptions in aviation propagate through financial markets.