eps-estimates-collector 0.2.5

Extract quarterly EPS estimates from FactSet Earnings Insight reports using OCR

EPS Estimates Collector

A Python package for extracting quarterly EPS (Earnings Per Share) estimates from financial reports using OCR and image processing techniques.

⚠️ Disclaimer: This package is for educational and research purposes only. For production use, please use FactSet's official API. This package processes publicly available PDF reports and is not affiliated with or endorsed by FactSet.

Overview

This project processes chart images containing S&P 500 quarterly EPS data and extracts quarter labels (e.g., Q1'14, Q2'15) and corresponding EPS values. The extracted data is saved in CSV format for further analysis.

Motivation

Financial data providers (FactSet, Bloomberg, Investing.com, etc.) typically offer historical EPS data as actual values—once a quarter's earnings are reported, the estimate is overwritten with the actual figure. This creates a challenge for backtesting predictive models: using historical data means testing against information that was already reflected in stock prices at the time, making it difficult to evaluate the true predictive power of EPS estimates.

To address this, this project extracts point-in-time EPS estimates from historical earnings insight reports. By preserving the estimates as they appeared at each report date (before actual earnings were announced), a dataset can be built that accurately reflects what was known and expected at each point in time, enabling more meaningful backtesting and predictive analysis.

Installation

Install from PyPI:

pip install eps-estimates-collector

Or with uv:

uv pip install eps-estimates-collector

Workflow Overview

The complete workflow from PDF documents to final P/E ratio calculation:

┌─────────────────────────────────────────────────────────────────────┐
│                    📄 Step 1: PDF Download                          │
│                                                                     │
│  FactSet Earnings Insight Reports                                   │
│  └─> Download PDFs from FactSet website                             │
│      (e.g., EarningsInsight_20251114_111425.pdf)                    │
└─────────────────────────────────────────────────────────────────────┘
                            │
                            ▼
┌─────────────────────────────────────────────────────────────────────┐
│              🖼️  Step 2: EPS Chart Page Extraction                  │
│                                                                     │
│  PDF Document                                                       │
│  └─> Extract EPS chart page (Page 6)                                │
│      └─> Convert to PNG image                                       │
│          (e.g., 20161209-6.png)                                     │
└─────────────────────────────────────────────────────────────────────┘
                            │
                            ▼
┌─────────────────────────────────────────────────────────────────────┐
│              🔍 Step 3: OCR Processing & Data Extraction            │
│                                                                     │
│  Chart Image                                                        │
│  ├─> Google Cloud Vision API (149 text regions detected)            │
│  ├─> Coordinate-based matching (Q1'14 ↔ 27.85)                      │
│  ├─> Bar classification (dark = actual, light = estimate)           │
│  └─> Extract quarter labels and EPS values                          │
│                                                                     │
│  Output: CSV with quarterly EPS estimates                           │
│  └─> extracted_estimates.csv                                        │
└─────────────────────────────────────────────────────────────────────┘
                            │
                            ▼
┌─────────────────────────────────────────────────────────────────────┐
│              📊 Step 4: P/E Ratio Calculation                       │
│                                                                     │
│  EPS Estimates + S&P 500 Prices                                     │
│  ├─> Load EPS data from public URL                                  │
│  ├─> Load S&P 500 prices from yfinance (2016-12-09 to today)        │
│  ├─> Calculate 4-quarter EPS sum (forward: Q(1)+Q(2)+Q(3)+Q(4), etc.) │
│  └─> Calculate P/E Ratio = Price / EPS_4Q_Sum                       │
│                                                                     │
│  Output: DataFrame with P/E ratios                                  │
└─────────────────────────────────────────────────────────────────────┘

Visual Workflow

Step 1: PDF Document → Downloads FactSet Earnings Insight PDF reports

Step 2: EPS Chart Page Extraction → Extracts chart page from PDF and converts to PNG image

Step 3: OCR Processing & Bar Classification → Extracts quarter labels and EPS values, classifies bars (dark = actual, light = estimate)

Step 4: P/E Ratio Calculation → See example output below

Usage

Python API

from eps_estimates_collector import calculate_pe_ratio

# Calculate P/E ratios (auto-loads CSV and S&P 500 prices)
pe_df = calculate_pe_ratio(type='forward')
print(pe_df)

P/E Types:

• forward: Q(1) + Q(2) + Q(3) + Q(4) - Next 4 quarters after report date (skips current quarter)
• mix: Q(0) + Q(1) + Q(2) + Q(3) - Current quarter + next 3 quarters
• trailing-like: Q(-3) + Q(-2) + Q(-1) + Q(0) - Last 3 quarters before + current quarter (note: current quarter is an estimate, so this is not exact TTM)

Example: P/E Ratio Calculation Result

from eps_estimates_collector import calculate_pe_ratio

# Calculate trailing-like P/E ratios
pe_df = calculate_pe_ratio('trailing-like')
print(pe_df)

Output:

📈 Loading S&P 500 price data from yfinance (2016-12-09 to 2025-11-20)...
✅ Loaded 2249 S&P 500 price points

      Report_Date  Price_Date    Price  EPS_4Q_Sum  PE_Ratio         Type
0      2016-12-09  2016-12-09  2249.69      122.28     18.40  trailing-like
1      2016-12-09  2016-12-12  2257.48      122.28     18.46  trailing-like
2      2016-12-09  2016-12-13  2271.72      122.28     18.58  trailing-like
...
2246   2025-11-07  2025-11-13  6600.00      278.30     23.72  trailing-like
2247   2025-11-14  2025-11-14  6700.00      278.84     24.03  trailing-like
2248   2025-11-14  2025-11-19  6700.00      278.84     24.03  trailing-like

[2249 rows x 6 columns]

API Reference

calculate_pe_ratio(type='forward')

Calculate P/E ratios from EPS estimates using S&P 500 prices.

Parameters:

• type (str): 'forward', 'mix', or 'trailing-like'
  • 'forward': Q(1) + Q(2) + Q(3) + Q(4) - Next 4 quarters after report date (skips current quarter)
  • 'mix': Q(0) + Q(1) + Q(2) + Q(3) - Report date quarter + next 3 quarters
  • 'trailing-like': Q(-3) + Q(-2) + Q(-1) + Q(0) - Last 3 quarters before report date + report date quarter (note: report date quarter is an estimate, so this is not exact TTM)

Returns: DataFrame with columns:

• Report_Date: EPS report date
• Price_Date: Trading day price date
• Price: S&P 500 closing price
• EPS_4Q_Sum: 4-quarter EPS sum
• PE_Ratio: Calculated P/E ratio
• Type: P/E type used

Features:

• ✅ No API keys required
• ✅ Always loads latest data from public URL
• ✅ No local files needed
• ✅ Auto-loads S&P 500 prices from yfinance

Legal Disclaimer

This package is provided for educational and research purposes only.

• This package processes publicly available PDF reports from FactSet's website
• The data extraction and processing methods are implemented for academic research
• This package is NOT affiliated with, endorsed by, or sponsored by FactSet
• For production use, please use FactSet's official API

No Warranty: This software is provided "as is" without warranty of any kind, express or implied, including but not limited to the warranties of merchantability, fitness for a particular purpose, and noninfringement.

Limitation of Liability: In no event shall the authors or copyright holders be liable for any claim, damages, or other liability arising from the use of this software.

Data Usage: Users are responsible for ensuring compliance with FactSet's terms of service and any applicable data usage agreements when using this package.

License

MIT License

Links

• GitHub: seung-gu/eps-estimates-collector
• PyPI: eps-estimates-collector