(책 요약) Quantitative Finance with Python (Chapter 6 - Working with Financial Datasets)
핵심 내용
- Data sources
- Yahoo Finance: stock price data + some futures data.
- FRED: historical data on economic variables (GDP, employment, credit spreads) (fredapi in python)
- Treasury.gov: Historical yield curve data for the US
- Quandl: Equity market data + data on futures
- HistData: Contains free data from FX markets, including intra-day data
- OptionMetrics: Contains relatively clean options and futures data for equity and other markets (great, but costly)
- CRSP: A broad, robust, historical database that doesn’t suffer from survivorship bias (e.g. Equity prices)
- Ken French’s Website: Useful historical datasets of the returns for Fama-French factors
- Missing data filling
- Interpolation & Filling Forward
- Filling via Regression
- Filling via Bootstrapping
- Filling via K-Nearest Neighbor
- Outlier detection
- Single vs. Multi-Variate Outlier Detection
- Plotting
- Standard Deviation
- Density Analysis
- Distance from K-Nearest Neighbor