qis.generate_multi_asset_factsheet

qis.generate_multi_asset_factsheet(prices, benchmark_prices=None, benchmark=None, add_benchmarks_to_navs=True, perf_params=PerfParams(freq='W-WED', freq_vol='W-WED', freq_skewness='ME', freq_drawdown='D', freq_reg='W-WED', freq_excess_return='W-WED', return_type=<ReturnTypes.LOG: 'Log'>, sharpe_convention=<SharpeConvention.PA: 1>, rates_data=None), regime_classifier=<qis.perfstats.regime_classifier.BenchmarkReturnsQuantilesRegime object>, heatmap_freq='YE', time_period=None, figsize=(8.3, 11.7), fontsize=5, factsheet_name=None, performance_bars=(PerfStat.Sharpe (rf=0), PerfStat.Max DD), drop_1y_ra_perf_table=True, min_trailing_obs=12, **kwargs)[source]

one-page cross-sectional factsheet comparing instruments against a benchmark.

The multi-asset archetype behind qis.factsheet(): cumulative performance with regime shading, a risk-adjusted performance table, rolling statistics, drawdowns, periodic returns and correlations, laid out on a single A4 page.

Parameters:

  • prices (DataFrame) – price panel, one column per instrument

  • benchmark_prices (Series | DataFrame) – benchmark panel; when given, its columns are the regression and regime reference

  • benchmark (str) – column name to use as the reference; defaults to the first benchmark column

  • add_benchmarks_to_navs (bool) – include the benchmarks as rows in the performance panels rather than only as the regression reference

  • perf_params (PerfParams) – sampling frequencies and Sharpe convention for every statistic

  • regime_classifier (BenchmarkReturnsQuantilesRegime) – how the benchmark history is cut into regimes for the conditional panels

  • heatmap_freq (str) – aggregation of the periodic-returns heatmap, ‘YE’ for calendar years

  • time_period (TimePeriod) – reporting span; defaults to the full history

  • figsize (Tuple[float, float]) – figure size in inches; the default is A4 portrait

  • fontsize (int) – base font size for the tables

  • factsheet_name (str) – report title

  • performance_bars (Tuple[PerfStat, PerfStat]) – the two statistics drawn as bar panels

  • drop_1y_ra_perf_table (bool) – omit the trailing one-year table, which is noise on a long history

  • min_trailing_obs (int) – minimum observations before a trailing statistic is reported

  • **kwargs – forwarded to the underlying plot functions

Returns:

the assembled figure

Return type:

Figure