finds.backtesting.riskpremium

Evaluate risk premiums from cross-sectional regressions

Copyright 2022, Terence Lim

MIT License

class finds.backtesting.riskpremium.RiskPremium(sql: SQL, bench: Benchmarks, rf: str, end: int)

Bases: object

Compute and test of factor loading risk premiums

Parameters:

• sql – Connection to user database to store results

• bench – Benchmark dataset of market and indexreturns

• rf – Name of riskfree rate from bench database

• end – Last date of price and returns data

__call__(stocks: Stocks, loadings: Dict[int, DataFrame], weights: str = '', standardize: List[str] = []) → Series

Estimate factor risk premiums with cross-sectional FM regressions

Parameters:

• stocks – Stocks’ returns data

• loadings – dict keyed by rebalance date of loadings DataFrames

• standardize – List of columns to demean and rescale (eql-wtd std = 1)

• weights – List of weights for weighted least squares and demean

Returns:

Series of means and stderrs of FM cross-sectional regression

fit(benchnames: List[str] = []) → List[DataFrame]

Compute risk premiums and benchmark correlations

plot(factors: List[str] = [], num: int = 1, fontsize: int = 8, figsize: Tuple[float, float] = (10, 5))

Plot computed time series of factor returns