image imagewidth (px) 1.28k 1.28k |
|---|
YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
Stock Volume and Return Analysis
1. Dataset Overview
This project analyzes stock market data from July 2025, using a dataset containing daily information such as open, close, high, and low prices, trading volume, and company fundamentals.
File used: stock_data_aug_2025.csv
Rows: 2,542
Columns: 14
Period covered: July 1–31, 2025
2. Research Question
The analysis focuses on identifying if increasing trading volume after monthly low ( in this case during July 2025) are correlated with greater return in stocks.
3. Methodology
- Loaded and cleaned the dataset — verified data types, removed duplicates, and filtered invalid prices.
- Focused on stocks that reached a monthly low during July.
- Calculated the average trading volume after that low point.
- Divided the stocks into two groups:
- High Volume after Low (above the monthly mean)
- Low Volume after Low (below the monthly mean)
- Computed the average stock return for 5 days, 10 days and 1 day following the low.
- Visualized and compared the results.
4. Results Summary
| Time Window | High Volume Avg Return | Low Volume Avg Return |
|---|---|---|
| 5 Days | 8.64% | 8.37% |
| 10 Days | 7.47% | 7.27% |
| 1 Day | 9.22% | 6.87% |
After examining the research question on the 5 days time frame, i wasn't satisfied with the results and decided to enlarge the time frame to 10 days in order to see if this will affect the results. The results whithin the 10 days time frame didn't show any conclusions eather, so i changed the time frame again to see if there is an affect for the short term after low. The results showed a 2.35 % difference between low and high volume, which is a significant difference in stocks trading.
5. Visualizations
5-Day Return by Volume Group
10-Day Return by Volume Group
1-Day Return by Volume Group
Combined Comparison of All Time Windows
6. Insights and Conclusions
The analysis shows that trading volume alone does not strongly predict price movement after a market low.
Even when extending the period from 5 to 10 trading days, the difference between high- and low-volume groups remains small.
However, when zooming in to a short-term period ( 1DAY time frame), the results show a slight correlation which can be helpfull as an indicator for examining monthly low stocks.
Although the tests showed some correlation regarding the research question, it is still not enough to prove the point and use the the correlation as a tool.
7. Tools Used
- Python (Pandas, Plotnine, NumPy)
- Google Colab
- Hugging Face Datasets
- Markdown for documentation
8. Author
Name: [Gal Solomon]
Course: EDA & Data Analysis Assignment
Platform: Google Colab / Hugging Face
Date: 16/11/2025
9. video
(https://www.loom.com/share/13d80041930646ac97d80ff35b3014a1)
- Downloads last month
- 7