A daily momentum screener for luxury watch market analysis that identifies hot watch references by combining real-time market data, price momentum, supply/demand dynamics, and profit calculations to help traders make informed buying decisions.
Daily screener for hot watch references (Rolex, Omega, Cartier, AP to start).
- Scrapes Chrono24 for current market prices and listing counts (free alternative to WatchCharts API @ $1000/yr)
- Pulls eBay Browse API counts as a demand proxy (optional, pending developer approval)
- Computes Δ7/14/30d, Z90, supply deltas (when available), and a composite Heat Score
- Adds Profit Overlay with Max Bid (8%) and Max Bid (10%) based on fees/buffers
Watch Heat is a comprehensive momentum screener designed for luxury watch market analysis. It automates the process of identifying "hot" watch references by aggregating data from multiple sources, computing sophisticated momentum metrics, and providing actionable profit calculations.
Key Capabilities:
- Real-time Market Data: Pulls current market prices, active listings, and days-on-market metrics from WatchCharts API
- Demand Signals: Tracks eBay activity as a proxy for market demand
- Momentum Analysis: Calculates multi-timeframe price changes (7/14/30 days), rolling z-scores, and supply/demand deltas
- Composite Heat Score: Weighted algorithm combining price momentum, market dynamics, and demand signals
- Profit Optimization: Calculates maximum bid prices to achieve target margins (8-10%) after accounting for fees and costs
- Interactive Reports: Beautiful, sortable HTML reports with filtering and search capabilities
Perfect for watch traders, collectors, and market analysts who want to spot emerging trends and make data-driven buying decisions.
- Automated Daily Screening: Fetches latest market data and computes momentum metrics
- Heat Score Calculation: Composite score combining price momentum, supply/demand dynamics, and market signals
- Profit Analysis: Calculates maximum bid prices to achieve target margins (8-10%)
- Interactive HTML Reports: Sortable, filterable reports with modern UI
- Robust Error Handling: Retry logic with exponential backoff for API calls
- Comprehensive Logging: Detailed logs saved to
watch_heat.log - Data Validation: Validates inputs and handles edge cases gracefully
- Flexible Configuration: Command-line arguments for custom output directories and dates
cd watch-heat
python -m venv .venv && source .venv/bin/activate
pip install -r requirements.txt
# Create .env file with your API keys (see .env.template)
python main.pypython main.py --help
Options:
--output-dir DIR Output directory for reports (default: data/)
--date YYYY-MM-DD Date to analyze (default: today or most recent available)
--verbose, -v Enable verbose loggingNo API key required! Data is scraped using Playwright browser automation.
Note: Chrono24 has aggressive Cloudflare protection. The scraper uses:
- Fresh browser contexts per request
- User agent rotation
- 5-second delays between requests
- Retry logic with exponential backoff
- Create an app at eBay Developers
- Wait for developer approval (~1 day)
- Generate an OAuth token using client credentials flow
- Add to
.env:EBAY_OAUTH_TOKEN=your_oauth_token
WatchCharts API costs $1000/year. We use Chrono24 scraping as a free alternative.
watch-heat/
├── main.py # Main entry point
├── settings.py # Configuration and API keys
├── universe.csv # Watch references to track
├── requirements.txt # Python dependencies
├── data_sources/
│ ├── chrono24_scraper.py # Chrono24 Playwright scraper (primary data source)
│ ├── watchcharts.py # Wrapper that delegates to chrono24_scraper
│ ├── watchcharts_client.py # Legacy WatchCharts API (not used - $1000/yr)
│ └── ebay.py # eBay Browse API client (optional)
├── analytics/
│ ├── metrics.py # Heat score and momentum calculations
│ └── profit.py # Profit overlay calculations
├── report/
│ └── render.py # HTML report generation
├── cache/ # Daily snapshot cache (auto-created)
└── data/ # Output reports (auto-created)
Edit settings.py to customize:
- MIN_LISTINGS: Minimum active listings to include (default: 5)
- HEAT_THRESHOLD: Heat score threshold for "hot" watches (default: 0.75)
- LOOKBACK_DAYS: Historical data window (default: 90)
- Profit Model: Fees, margins, shipping costs (see
TARGET_MARGIN_*,*_FEE_RATE, etc.)
The heat score is a weighted composite of:
- 14-day price change (35%): Short-term momentum
- 30-day price change (25%): Medium-term trend
- DOM delta (20%): Days on market change (negative = faster sales = positive)
- Supply delta (20%): Active listings change (negative = lower supply = positive)
- Z90 score (10%): 90-day rolling z-score (capped at 3.0)
- eBay momentum (10%): 30-day normalized eBay activity change
The script generates two files in the data/ directory:
- CSV file (
watch_heat_YYYY-MM-DD.csv): Raw data for analysis - HTML file (
watch_heat_YYYY-MM-DD.html): Interactive report with:- Summary statistics
- Sortable columns
- Search/filter functionality
- Color-coded heat scores
- Hot watch highlighting
- The pipeline persists daily snapshots in
cache/. Heat metrics become meaningful after 7-30 days of runs. - Chrono24 scraping takes ~15-20 seconds per watch due to Cloudflare protection and rate limiting.
- The script includes automatic retry logic (3 attempts) for scraping.
- All operations are logged to
watch_heat.logfor debugging. - Missing data is handled gracefully with appropriate warnings in logs.
- DOM (days on market) data is not available from Chrono24 - this metric will show as blank.
- ✅ Switched to Chrono24 scraping - Free alternative to WatchCharts ($1000/yr saved!)
- ✅ Playwright-based scraper with Cloudflare bypass
- ✅ User agent rotation and fresh browser contexts
- ✅ Comprehensive logging and error handling
- ✅ Retry logic with delays for rate limiting
- ✅ Enhanced HTML report with sorting, filtering, and modern UI
- ✅ Command-line arguments for flexibility
- ✅ Data validation throughout the pipeline