In this project, I developed a quantitative regression model in SAS to analyze insider trading activities. The model utilized inputs such as analyst forecast errors, accruals, asset turnover, quick ratio, and market-adjusted returns to identify significant trading patterns and potential instances of insider trading.
Identifying Insider Trading Part 1.pdf
Identifying Insider Trading Part 2.pdf
Identifying Insider Trading Part 3.pdf
Identifying Insider Trading Presentation.pdf
<aside> 💡 How might we leverage quantitative data analysis to detect significant patterns and identify potential instances of insider trading?
</aside>
Why we chose to address this problem:
The primary reason for our investigation was the apparent conflict of interest for internal executives who have access to insider information and engage in trading activities. This issue raises significant concerns about market integrity and fairness. Our objective is to develop a reproducible and systematic method to identify such insider trading practices, ultimately enhancing transparency, regulatory compliance, and protecting the interests of all market participants.
In Part 1, we began by estimating analyst forecast errors using a model that incorporated variables such as accruals, asset turnover, quick ratio, and operating margin. This foundational work established the critical predictors for our regression analysis, enabling us to understand how these variables influenced the accuracy of analyst forecasts.
Building on this, Part 2 focused on portfolio sorts based on predicted forecast errors. By sorting firms into quintiles and evaluating their market-adjusted returns over a 360-day holding period, we demonstrated the model's predictive power and its ability to identify profitable trading strategies. This step was crucial in validating the effectiveness of our chosen variables and regression model.
In Part 3, we aggregated individual trading data to calculate key statistics such as the number of buys and sells, total dollar volume traded, and market-adjusted returns over 180 days. This allowed us to create detailed histograms and identify outliers, further refining our model by focusing on individuals with significant trading activity and performance.
Finally, our comprehensive final presentation synthesized these findings, showcasing the model's ability to uncover patterns of insider trading. The presentation highlighted notable cases and provided a clear visual representation of our results, demonstrating how the regression model systematically identified potential insider trading activities.
Through this multi-faceted approach, leveraging regression analysis and a detailed examination of insider trading patterns, our solution offers a reproducible method to detect and address conflicts of interest among internal executives. This model not only enhances transparency and regulatory compliance but also provides a powerful tool for market participants and regulators to ensure fair trading practices.
See the documents below.
Identifying Insider Trading Part 1.pdf
Identifying Insider Trading Part 2.pdf