OpenFrame 

all_properties(properties=None)[source]

Calculate chosen timeseries properties.

Parameters:

properties (list[Literal['value_ret', 'geo_ret', 'arithmetic_ret', 'vol', 'downside_deviation', 'ret_vol_ratio', 'sortino_ratio', 'kappa3_ratio', 'z_score', 'skew', 'kurtosis', 'positive_share', 'var_down', 'cvar_down', 'vol_from_var', 'worst', 'worst_month', 'max_drawdown', 'max_drawdown_date', 'max_drawdown_cal_year', 'first_indices', 'last_indices', 'lengths_of_items', 'span_of_days_all']] | None) – The properties to calculate. Defaults to calculating all available. Optional.
self (Self)

Returns:

Properties of the constituent OpenTimeSeries.

Return type:

property lengths_of_items: Series[int]

Number of observations of all constituents.

Returns:: Number of observations of all constituents.

property item_count: int

Number of constituents.

Returns:: Number of constituents.

property columns_lvl_zero: list[str]

Level 0 values of the MultiIndex columns in the .tsdf DataFrame.

Returns:: Level 0 values of the MultiIndex columns in the .tsdf DataFrame.

property columns_lvl_one: list[ValueType]

Level 1 values of the MultiIndex columns in the .tsdf DataFrame.

Returns:: Level 1 values of the MultiIndex columns in the .tsdf DataFrame.

property first_indices: Series[dt.date]

The first dates in the timeseries of all constituents.

Returns:: The first dates in the timeseries of all constituents.

property last_indices: Series[dt.date]

The last dates in the timeseries of all constituents.

Returns:: The last dates in the timeseries of all constituents.

property span_of_days_all: Series[int]

Number of days from the first date to the last for all items in the frame.

Returns:: Number of days from the first date to the last for all items in the frame.

value_to_ret()[source]

Convert series of values into series of returns.

Returns:: The returns of the values in the series.
Parameters:: self (Self)
Return type:: Self

value_to_diff(periods=1)[source]

Convert series of values to series of their period differences.

Parameters:

periods (int) – The number of periods between observations over which difference is calculated. Defaults to 1.
self (Self)

Returns:

An OpenFrame object.

Return type:

to_cumret()[source]

Convert series of returns into cumulative series of values.

Returns:: An OpenFrame object.
Parameters:: self (Self)
Return type:: Self

resample(freq='BME')[source]

Resample the timeseries frequency.

Parameters:

freq (Literal['B', 'BME', 'BQE', 'BYE'] | str) – The date offset string that sets the resampled frequency. Defaults to “BME”.
self (Self)

Returns:

An OpenFrame object.

Return type:

resample_to_business_period_ends(freq='BME', method='nearest')[source]

Resamples timeseries frequency to the business calendar month end dates.

Stubs left in place. Stubs will be aligned to the shortest stub.

Parameters:

freq (Literal['B', 'BME', 'BQE', 'BYE']) – The date offset string that sets the resampled frequency. Defaults to “BME”.
method (Literal['pad', 'ffill', 'backfill', 'bfill', 'nearest'] | None) – Controls the method used to align values across columns. Defaults to nearest.
self (Self)

Returns:

An OpenFrame object.

Return type:

ewma_risk(lmbda=0.94, day_chunk=11, dlta_degr_freedms=0, first_column=0, second_column=1, corr_scale=2.0, months_from_last=None, from_date=None, to_date=None, periods_in_a_year_fixed=None)[source]

Exponentially Weighted Moving Average Volatilities and Correlation.

Exponentially Weighted Moving Average (EWMA) for Volatilities and Correlation.

Reference: https://www.investopedia.com/articles/07/ewma.asp.

Parameters:

lmbda (float) – Scaling factor to determine weighting. Defaults to 0.94.
day_chunk (int) – Sampling the data which is assumed to be daily. Defaults to 11.
dlta_degr_freedms (int) – Variance bias factor taking the value 0 or 1. Defaults to 0.
first_column (int) – Column of first timeseries. Defaults to 0.
second_column (int) – Column of second timeseries. Defaults to 1.
corr_scale (float) – Correlation scale factor. Defaults to 2.0.
months_from_last (int | None) – Number of months offset as positive integer. Overrides use of from_date and to_date. Optional.
from_date (dt.date | None) – Specific from date. Optional.
to_date (dt.date | None) – Specific to date. Optional.
periods_in_a_year_fixed (DaysInYearType | None) – Allows locking the periods-in-a-year to simplify test cases and comparisons. Optional.
self (Self)

Returns:

Series volatilities and correlation.

Return type:

DataFrame

property correl_matrix: DataFrame

Correlation matrix.

This property returns the correlation matrix of the time series in the frame.

Returns:: Correlation matrix of the time series in the frame.

add_timeseries(new_series)[source]

To add an OpenTimeSeries object.

Parameters:

new_series (OpenTimeSeries) – The timeseries to add.
self (Self)

Returns:

An OpenFrame object.

Return type:

delete_timeseries(lvl_zero_item)[source]

To delete an OpenTimeSeries object.

Parameters:

lvl_zero_item (str) – The .tsdf column level 0 value of the timeseries to delete.
self (Self)

Returns:

An OpenFrame object.

Return type:

trunc_frame(start_cut=None, end_cut=None, where='both')[source]

Truncate DataFrame such that all timeseries have the same time span.

Parameters:

start_cut (dt.date | None) – New first date. Optional.
end_cut (dt.date | None) – New last date. Optional.
where (LiteralTrunc) – Determines where dataframe is truncated also when start_cut or end_cut is None. Defaults to both.
self (Self)

Returns:

An OpenFrame object.

Return type:

Self

relative(long_column=0, short_column=1, *, base_zero=True)[source]

Calculate cumulative relative return between two series.

Parameters:

long_column (int) – Column number of timeseries bought. Defaults to 0.
short_column (int) – Column number of timeseries sold. Defaults to 1.
base_zero (bool) – If set to False 1.0 is added to allow for a capital base and to allow a volatility calculation. Defaults to True.
self (Self)

Return type:

None

tracking_error_func(base_column=-1, months_from_last=None, from_date=None, to_date=None, periods_in_a_year_fixed=None)[source]

Tracking Error.

Calculates Tracking Error which is the standard deviation of the difference between the fund and its index returns.

Reference: https://www.investopedia.com/terms/t/trackingerror.asp.

Parameters:

base_column (tuple[str, ValueType] | int) – Column of timeseries that is the denominator in the ratio. Defaults to -1.
months_from_last (int | None) – Number of months offset as positive integer. Overrides use of from_date and to_date. Optional.
from_date (dt.date | None) – Specific from date. Optional.
to_date (dt.date | None) – Specific to date. Optional.
periods_in_a_year_fixed (DaysInYearType | None) – Allows locking the periods-in-a-year to simplify test cases and comparisons. Optional.
self (Self)

Returns:

Tracking Errors.

Return type:

Series[float]

info_ratio_func(base_column=-1, months_from_last=None, from_date=None, to_date=None, periods_in_a_year_fixed=None)[source]

Information Ratio.

The Information Ratio equals ( fund return less index return ) divided by the Tracking Error. And the Tracking Error is the standard deviation of the difference between the fund and its index returns. The ratio is calculated using the annualized arithmetic mean of returns.

Parameters:

base_column (tuple[str, ValueType] | int) – Column of timeseries that is the denominator in the ratio. Defaults to -1.
months_from_last (int | None) – Number of months offset as positive integer. Overrides use of from_date and to_date. Optional.
from_date (dt.date | None) – Specific from date. Optional.
to_date (dt.date | None) – Specific to date. Optional.
periods_in_a_year_fixed (DaysInYearType | None) – Allows locking the periods-in-a-year to simplify test cases and comparisons. Optional.
self (Self)

Returns:

Information Ratios.

Return type:

Series[float]

capture_ratio_func(ratio, base_column=-1, months_from_last=None, from_date=None, to_date=None, periods_in_a_year_fixed=None)[source]

Capture Ratio.

The Up (Down) Capture Ratio is calculated by dividing the CAGR of the asset during periods that the benchmark returns are positive (negative) by the CAGR of the benchmark during the same periods. CaptureRatio.BOTH is the Up ratio divided by the Down ratio. Source: ‘Capture Ratios: A Popular Method of Measuring Portfolio Performance in Practice’, Don R. Cox and Delbert C. Goff, Journal of Economics and Finance Education (Vol 2 Winter 2013).

Reference: https://www.economics-finance.org/jefe/volume12-2/11ArticleCox.pdf.

Parameters:

ratio (LiteralCaptureRatio) – The ratio to calculate.
base_column (tuple[str, ValueType] | int) – Column of timeseries that is the denominator in the ratio. Defaults to -1.
months_from_last (int | None) – Number of months offset as positive integer. Overrides use of from_date and to_date. Optional.
from_date (dt.date | None) – Specific from date. Optional.
to_date (dt.date | None) – Specific to date. Optional.
periods_in_a_year_fixed (DaysInYearType | None) – Allows locking the periods-in-a-year to simplify test cases and comparisons. Optional.
self (Self)

Returns:

Capture Ratios.

Return type:

Series[float]

beta(asset, market, dlta_degr_freedms=1)[source]

Market Beta.

Calculates Beta as Co-variance of asset & market divided by Variance of the market.

Reference: https://www.investopedia.com/terms/b/beta.asp.

Parameters:

asset (tuple[str, ValueType] | int) – The column of the asset.
market (tuple[str, ValueType] | int) – The column of the market against which Beta is measured.
dlta_degr_freedms (int) – Variance bias factor taking the value 0 or 1. Defaults to 1.
self (Self)

Returns:

Beta as Co-variance of x & y divided by Variance of x.

Return type:

ord_least_squares_fit(y_column, x_column, *, fitted_series=True)[source]

Ordinary Least Squares fit.

Performs a linear regression and adds a new column with a fitted line using Ordinary Least Squares fit.

Parameters:

y_column (tuple[str, ValueType] | int) – The column level values of the dependent variable y.
x_column (tuple[str, ValueType] | int) – The column level values of the exogenous variable x.
fitted_series (bool) – If True the fit is added as a new column in the .tsdf Pandas.DataFrame. Defaults to True.
self (Self)

Returns:

A dictionary with the coefficient, intercept and rsquared outputs.

Return type:

dict[str, float]

jensen_alpha(asset, market, riskfree_rate=0.0, dlta_degr_freedms=1)[source]

Jensen’s alpha.

The Jensen’s measure, or Jensen’s alpha, is a risk-adjusted performance measure that represents the average return on a portfolio or investment, above or below that predicted by the capital asset pricing model (CAPM), given the portfolio’s or investment’s beta and the average market return. This metric is also commonly referred to as simply alpha.

Reference: https://www.investopedia.com/terms/j/jensensmeasure.asp.

Parameters:

asset (tuple[str, ValueType] | int) – The column of the asset.
market (tuple[str, ValueType] | int) – The column of the market against which Jensen’s alpha is measured.
riskfree_rate (float) – The return of the zero volatility riskfree asset. Defaults to 0.0.
dlta_degr_freedms (int) – Variance bias factor taking the value 0 or 1. Defaults to 1.
self (Self)

Returns:

Jensen’s alpha.

Return type:

make_portfolio(name, weight_strat=None)[source]

Calculate a basket timeseries based on the supplied weights.

Parameters:

name (str) – Name of the basket timeseries.
weight_strat (Literal['eq_weights', 'inv_vol', 'max_div', 'min_vol_overweight'] | None) – Weight calculation strategies. Optional.
self (Self)

Returns:

A basket timeseries.

Return type:

rolling_info_ratio(long_column=0, short_column=1, observations=21, periods_in_a_year_fixed=None)[source]

Calculate rolling Information Ratio.

The Information Ratio equals ( fund return less index return ) divided by the Tracking Error. And the Tracking Error is the standard deviation of the difference between the fund and its index returns.

Parameters:

long_column (int) – Column of timeseries that is the numerator in the ratio. Defaults to 0.
short_column (int) – Column of timeseries that is the denominator in the ratio. Defaults to 1.
observations (int) – The length of the rolling window to use is set as number of observations. Defaults to 21.
periods_in_a_year_fixed (Annotated[int, FieldInfo(annotation=NoneType, required=True, metadata=[Strict(strict=True), Ge(ge=1), Le(le=366)])] | None) – Allows locking the periods-in-a-year to simplify test cases and comparisons. Optional.
self (Self)

Returns:

Rolling Information Ratios.

Return type:

rolling_beta(asset_column=0, market_column=1, observations=21, dlta_degr_freedms=1)[source]

Calculate rolling Market Beta.

Calculates Beta as Co-variance of asset & market divided by Variance of the market.

Reference: https://www.investopedia.com/terms/b/beta.asp.

Parameters:

asset_column (int) – Column of timeseries that is the asset. Defaults to 0.
market_column (int) – Column of timeseries that is the market. Defaults to 1.
observations (int) – The length of the rolling window to use is set as number of observations. Defaults to 21.
dlta_degr_freedms (int) – Variance bias factor taking the value 0 or 1. Defaults to 1.
self (Self)

Returns:

Rolling Betas.

Return type:

rolling_corr(first_column=0, second_column=1, observations=21)[source]

Calculate rolling Correlation.

Calculates correlation between two series. The period with at least the given number of observations is the first period calculated.

Parameters:

first_column (int) – The position as integer of the first timeseries to compare. Defaults to 0.
second_column (int) – The position as integer of the second timeseries to compare. Defaults to 1.
observations (int) – The length of the rolling window to use is set as number of observations. Defaults to 21.
self (Self)

Returns:

Rolling Correlations.

Return type:

multi_factor_linear_regression(dependent_column)[source]

Perform a multi-factor linear regression.

This function treats one specified column in the DataFrame as the dependent variable (y) and uses all remaining columns as independent variables (X). It utilizes a scikit-learn LinearRegression model and returns a DataFrame with summary output and an OpenTimeSeries of predicted values.

Parameters:

dependent_column (tuple[str, ValueType]) – A tuple key to select the column in the OpenFrame.tsdf.columns to use as the dependent variable.
self (Self)

Returns:

A DataFrame with the R-squared, the intercept and the regression coefficients
An OpenTimeSeries of predicted values

Return type:

A tuple containing

Raises:

KeyError – If the column tuple is not found in the OpenFrame.tsdf.columns.
ValueError – If not all series are returnseries (ValueType.RTRN).

model_config = {'arbitrary_types_allowed': True, 'revalidate_instances': 'always', 'validate_assignment': True}: Configuration for the model, should be a dictionary conforming to [ConfigDict][pydantic.config.ConfigDict].

rebalanced_portfolio(name, items=None, bal_weights=None, frequency=1, cash_index=None, *, equal_weights=False, drop_extras=True)[source]

Create a rebalanced portfolio from the OpenFrame constituents.

Parameters:

name (str) – Name of the portfolio.
items (list[str] | None) – List of items to include in the portfolio. If None, uses all items. Optional.
bal_weights (list[float] | None) – List of weights for rebalancing. If None, uses frame weights. Optional.
frequency (int) – Rebalancing frequency. Defaults to 1.
cash_index (OpenTimeSeries | None) – Cash index series for cash component. Optional.
equal_weights (bool) – If True, use equal weights for all items. Defaults to False.
drop_extras (bool) – If True, only return TWR series; if False, return all details. Defaults to True.
self (Self)

Returns:

OpenFrame containing the rebalanced portfolio.

Return type:

OpenFrame

The OpenFrame class manages collections of OpenTimeSeries objects and provides functionality for:

Multi-asset analysis and comparison
Portfolio construction and optimization
Correlation and regression analysis
Risk attribution and factor analysis
Batch processing of multiple time series

Class Methods for Construction

OpenFrame.from_deepcopy()[source]

Create copy of the OpenFrame object.

Returns:: An OpenFrame object.
Parameters:: self (Self)
Return type:: Self

Properties

Frame-specific Properties

OpenFrame.constituents: list[OpenTimeSeries]

OpenFrame.columns_lvl_zero

Level 0 values of the MultiIndex columns in the .tsdf DataFrame.

Returns:: Level 0 values of the MultiIndex columns in the .tsdf DataFrame.

OpenFrame.columns_lvl_one

Level 1 values of the MultiIndex columns in the .tsdf DataFrame.

Returns:: Level 1 values of the MultiIndex columns in the .tsdf DataFrame.

OpenFrame.item_count

Number of constituents.

Returns:: Number of constituents.

OpenFrame.weights: list[float] | None

OpenFrame.first_indices

The first dates in the timeseries of all constituents.

Returns:: The first dates in the timeseries of all constituents.

OpenFrame.last_indices

The last dates in the timeseries of all constituents.

Returns:: The last dates in the timeseries of all constituents.

OpenFrame.lengths_of_items

Number of observations of all constituents.

Returns:: Number of observations of all constituents.

OpenFrame.span_of_days_all

Number of days from the first date to the last for all items in the frame.

Returns:: Number of days from the first date to the last for all items in the frame.

Common Properties

OpenFrame.first_idx

The first date in the timeseries.

Returns:: The first date in the timeseries.

OpenFrame.last_idx

The last date in the timeseries.

Returns:: The last date in the timeseries.

OpenFrame.length

Number of observations.

Returns:: Number of observations.

OpenFrame.span_of_days

Number of days from the first date to the last.

Returns:: Number of days from the first date to the last.

OpenFrame.tsdf: DataFrame

OpenFrame.max_drawdown_date

Date when the maximum drawdown occurred.

Reference: https://www.investopedia.com/terms/m/maximum-drawdown-mdd.asp.

Returns:

datetime.date | pandas.Series[dt.date]: Date when the maximum drawdown occurred

OpenFrame.periods_in_a_year

The average number of observations per year.

Returns:: The average number of observations per year.

OpenFrame.yearfrac

Length of series in years assuming 365.25 days per year.

Returns:: Length of the timeseries in years assuming 365.25 days per year.

Financial Metrics

OpenFrame.all_properties = <function OpenFrame.all_properties>[source]

Parameters:

self (Self)
properties (list[Literal['value_ret', 'geo_ret', 'arithmetic_ret', 'vol', 'downside_deviation', 'ret_vol_ratio', 'sortino_ratio', 'kappa3_ratio', 'z_score', 'skew', 'kurtosis', 'positive_share', 'var_down', 'cvar_down', 'vol_from_var', 'worst', 'worst_month', 'max_drawdown', 'max_drawdown_date', 'max_drawdown_cal_year', 'first_indices', 'last_indices', 'lengths_of_items', 'span_of_days_all']] | None)

Return type:

OpenFrame.arithmetic_ret

Annualized arithmetic mean of returns.

Reference: https://www.investopedia.com/terms/a/arithmeticmean.asp.

Returns:

SeriesOrFloat_co: Annualized arithmetic mean of returns. Returns float for OpenTimeSeries, Series[float] for OpenFrame.

OpenFrame.geo_ret

Compounded Annual Growth Rate (CAGR).

Reference: https://www.investopedia.com/terms/c/cagr.asp.

Returns:

SeriesOrFloat_co: Compounded Annual Growth Rate (CAGR). Returns float for OpenTimeSeries, Series[float] for OpenFrame.

OpenFrame.value_ret

Simple return.

Returns:

SeriesOrFloat_co: Simple return. Returns float for OpenTimeSeries, Series[float] for OpenFrame.

OpenFrame.vol

Annualized volatility.

Based on Pandas .std() which is the equivalent of stdev.s([…]) in MS Excel.

Reference: https://www.investopedia.com/terms/v/volatility.asp.

Returns:

SeriesOrFloat_co: Annualized volatility. Returns float for OpenTimeSeries, Series[float] for OpenFrame.

OpenFrame.downside_deviation

Downside Deviation.

Standard deviation of returns that are below a Minimum Accepted Return of zero. It is used to calculate the Sortino Ratio.

Reference: https://www.investopedia.com/terms/d/downside-deviation.asp.

Returns:

SeriesOrFloat_co: Downside deviation. Returns float for OpenTimeSeries, Series[float] for OpenFrame.

OpenFrame.ret_vol_ratio

Ratio of annualized arithmetic mean of returns and annualized volatility.

Returns:

SeriesOrFloat_co: Ratio of the annualized arithmetic mean of returns and annualized volatility. Returns float for OpenTimeSeries, Series[float] for OpenFrame.

OpenFrame.sortino_ratio

Sortino ratio.

Reference: https://www.investopedia.com/terms/s/sortinoratio.asp.

Returns:

SeriesOrFloat_co: Sortino ratio calculated as the annualized arithmetic mean of returns / downside deviation. The ratio implies that the riskfree asset has zero volatility, and a minimum acceptable return of zero. Returns float for OpenTimeSeries, Series[float] for OpenFrame.

OpenFrame.kappa3_ratio

Kappa-3 ratio.

The Kappa-3 ratio is a generalized downside-risk ratio defined as annualized arithmetic return divided by the cubic-root of the lower partial moment of order 3 (with respect to a minimum acceptable return, MAR). It penalizes larger downside outcomes more heavily than the Sortino ratio (which uses order 2).

Returns:

SeriesOrFloat_co: Kappa-3 ratio calculation with the riskfree rate and. Minimum Acceptable Return (MAR) both set to zero. Returns float for OpenTimeSeries, Series[float] for OpenFrame.

OpenFrame.omega_ratio

Omega ratio.

Reference: https://en.wikipedia.org/wiki/Omega_ratio.

Returns:

SeriesOrFloat_co: Omega ratio calculation. Returns float for OpenTimeSeries, Series[float] for OpenFrame.

OpenFrame.var_down

Downside 95% Value At Risk (VaR).

The equivalent of percentile.inc([…], 1-level) over returns in MS Excel. https://www.investopedia.com/terms/v/var.asp.

Returns:

SeriesOrFloat_co: Downside 95% Value At Risk (VaR). Returns float for OpenTimeSeries, Series[float] for OpenFrame.

OpenFrame.cvar_down

Downside 95% Conditional Value At Risk “CVaR”.

Reference: https://www.investopedia.com/terms/c/conditional_value_at_risk.asp.

Returns:

SeriesOrFloat_co: Downside 95% Conditional Value At Risk “CVaR”. Returns float for OpenTimeSeries, Series[float] for OpenFrame.

OpenFrame.worst

Most negative percentage change.

Returns:

SeriesOrFloat_co: Most negative percentage change. Returns float for OpenTimeSeries, Series[float] for OpenFrame.

OpenFrame.worst_month

Most negative month.

Returns:

SeriesOrFloat_co: Most negative month. Returns float for OpenTimeSeries, Series[float] for OpenFrame.

OpenFrame.max_drawdown

Maximum drawdown without any limit on date range.

Reference: https://www.investopedia.com/terms/m/maximum-drawdown-mdd.asp.

Returns:

SeriesOrFloat_co: Maximum drawdown without any limit on date range. Returns float for OpenTimeSeries, Series[float] for OpenFrame.

OpenFrame.max_drawdown_cal_year

Maximum drawdown in a single calendar year.

Reference: https://www.investopedia.com/terms/m/maximum-drawdown-mdd.asp.

Returns:

SeriesOrFloat_co: Maximum drawdown in a single calendar year. Returns float for OpenTimeSeries, Series[float] for OpenFrame.

OpenFrame.positive_share

The share of percentage changes that are greater than zero.

Returns:

SeriesOrFloat_co: The share of percentage changes that are greater than zero. Returns float for OpenTimeSeries, Series[float] for OpenFrame.

OpenFrame.vol_from_var

Implied annualized volatility from Downside 95% Value at Risk.

Assumes that returns are normally distributed.

Returns:

SeriesOrFloat_co: Implied annualized volatility from the Downside 95% VaR using the assumption that returns are normally distributed. Returns float for OpenTimeSeries, Series[float] for OpenFrame.

OpenFrame.skew

Skew of the return distribution.

Reference: https://www.investopedia.com/terms/s/skewness.asp.

Returns:

SeriesOrFloat_co: Skew of the return distribution. Returns float for OpenTimeSeries, Series[float] for OpenFrame.

OpenFrame.kurtosis

Kurtosis of the return distribution.

Reference: https://www.investopedia.com/terms/k/kurtosis.asp.

Returns:

SeriesOrFloat_co: Kurtosis of the return distribution. Returns float for OpenTimeSeries, Series[float] for OpenFrame.

OpenFrame.z_score

Z-score.

Reference: https://www.investopedia.com/terms/z/zscore.asp.

Returns:

SeriesOrFloat_co: Z-score as (last return - mean return) / standard deviation of returns. Returns float for OpenTimeSeries, Series[float] for OpenFrame.

Methods

Frame Management

OpenFrame.merge_series(how='outer')[source]

Merge index of Pandas Dataframes of the constituent OpenTimeSeries.

Parameters:

how (Literal['outer', 'inner']) – The Pandas merge method. Defaults to “outer”.
self (Self)

Returns:

An OpenFrame object.

Return type:

OpenFrame.trunc_frame(start_cut=None, end_cut=None, where='both')[source]

Truncate DataFrame such that all timeseries have the same time span.

Parameters:

start_cut (dt.date | None) – New first date. Optional.
end_cut (dt.date | None) – New last date. Optional.
where (LiteralTrunc) – Determines where dataframe is truncated also when start_cut or end_cut is None. Defaults to both.
self (Self)

Returns:

An OpenFrame object.

Return type:

Self

OpenFrame.add_timeseries(new_series)[source]

To add an OpenTimeSeries object.

Parameters:

new_series (OpenTimeSeries) – The timeseries to add.
self (Self)

Returns:

An OpenFrame object.

Return type:

OpenFrame.delete_timeseries(lvl_zero_item)[source]

To delete an OpenTimeSeries object.

Parameters:

lvl_zero_item (str) – The .tsdf column level 0 value of the timeseries to delete.
self (Self)

Returns:

An OpenFrame object.

Return type:

Portfolio Analysis

OpenFrame.relative(long_column=0, short_column=1, *, base_zero=True)[source]

Calculate cumulative relative return between two series.

Parameters:

long_column (int) – Column number of timeseries bought. Defaults to 0.
short_column (int) – Column number of timeseries sold. Defaults to 1.
base_zero (bool) – If set to False 1.0 is added to allow for a capital base and to allow a volatility calculation. Defaults to True.
self (Self)

Return type:

None

OpenFrame.make_portfolio(name, weight_strat=None)[source]

Calculate a basket timeseries based on the supplied weights.

Parameters:

name (str) – Name of the basket timeseries.
weight_strat (Literal['eq_weights', 'inv_vol', 'max_div', 'min_vol_overweight'] | None) – Weight calculation strategies. Optional.
self (Self)

Returns:

A basket timeseries.

Return type:

OpenFrame.rebalanced_portfolio(name, items=None, bal_weights=None, frequency=1, cash_index=None, *, equal_weights=False, drop_extras=True)[source]

Create a rebalanced portfolio from the OpenFrame constituents.

Parameters:

name (str) – Name of the portfolio.
items (list[str] | None) – List of items to include in the portfolio. If None, uses all items. Optional.
bal_weights (list[float] | None) – List of weights for rebalancing. If None, uses frame weights. Optional.
frequency (int) – Rebalancing frequency. Defaults to 1.
cash_index (OpenTimeSeries | None) – Cash index series for cash component. Optional.
equal_weights (bool) – If True, use equal weights for all items. Defaults to False.
drop_extras (bool) – If True, only return TWR series; if False, return all details. Defaults to True.
self (Self)

Returns:

OpenFrame containing the rebalanced portfolio.

Return type:

OpenFrame

Statistical Analysis

OpenFrame.ord_least_squares_fit(y_column, x_column, *, fitted_series=True)[source]

Ordinary Least Squares fit.

Performs a linear regression and adds a new column with a fitted line using Ordinary Least Squares fit.

Parameters:

y_column (tuple[str, ValueType] | int) – The column level values of the dependent variable y.
x_column (tuple[str, ValueType] | int) – The column level values of the exogenous variable x.
fitted_series (bool) – If True the fit is added as a new column in the .tsdf Pandas.DataFrame. Defaults to True.
self (Self)

Returns:

A dictionary with the coefficient, intercept and rsquared outputs.

Return type:

dict[str, float]

OpenFrame.beta(asset, market, dlta_degr_freedms=1)[source]

Market Beta.

Calculates Beta as Co-variance of asset & market divided by Variance of the market.

Reference: https://www.investopedia.com/terms/b/beta.asp.

Parameters:

asset (tuple[str, ValueType] | int) – The column of the asset.
market (tuple[str, ValueType] | int) – The column of the market against which Beta is measured.
dlta_degr_freedms (int) – Variance bias factor taking the value 0 or 1. Defaults to 1.
self (Self)

Returns:

Beta as Co-variance of x & y divided by Variance of x.

Return type:

OpenFrame.jensen_alpha(asset, market, riskfree_rate=0.0, dlta_degr_freedms=1)[source]

Jensen’s alpha.

The Jensen’s measure, or Jensen’s alpha, is a risk-adjusted performance measure that represents the average return on a portfolio or investment, above or below that predicted by the capital asset pricing model (CAPM), given the portfolio’s or investment’s beta and the average market return. This metric is also commonly referred to as simply alpha.

Reference: https://www.investopedia.com/terms/j/jensensmeasure.asp.

Parameters:

asset (tuple[str, ValueType] | int) – The column of the asset.
market (tuple[str, ValueType] | int) – The column of the market against which Jensen’s alpha is measured.
riskfree_rate (float) – The return of the zero volatility riskfree asset. Defaults to 0.0.
dlta_degr_freedms (int) – Variance bias factor taking the value 0 or 1. Defaults to 1.
self (Self)

Returns:

Jensen’s alpha.

Return type:

OpenFrame.tracking_error_func(base_column=-1, months_from_last=None, from_date=None, to_date=None, periods_in_a_year_fixed=None)[source]

Tracking Error.

Calculates Tracking Error which is the standard deviation of the difference between the fund and its index returns.

Reference: https://www.investopedia.com/terms/t/trackingerror.asp.

Parameters:

base_column (tuple[str, ValueType] | int) – Column of timeseries that is the denominator in the ratio. Defaults to -1.
months_from_last (int | None) – Number of months offset as positive integer. Overrides use of from_date and to_date. Optional.
from_date (dt.date | None) – Specific from date. Optional.
to_date (dt.date | None) – Specific to date. Optional.
periods_in_a_year_fixed (DaysInYearType | None) – Allows locking the periods-in-a-year to simplify test cases and comparisons. Optional.
self (Self)

Returns:

Tracking Errors.

Return type:

Series[float]

OpenFrame.info_ratio_func(base_column=-1, months_from_last=None, from_date=None, to_date=None, periods_in_a_year_fixed=None)[source]

Information Ratio.

The Information Ratio equals ( fund return less index return ) divided by the Tracking Error. And the Tracking Error is the standard deviation of the difference between the fund and its index returns. The ratio is calculated using the annualized arithmetic mean of returns.

Parameters:

base_column (tuple[str, ValueType] | int) – Column of timeseries that is the denominator in the ratio. Defaults to -1.
months_from_last (int | None) – Number of months offset as positive integer. Overrides use of from_date and to_date. Optional.
from_date (dt.date | None) – Specific from date. Optional.
to_date (dt.date | None) – Specific to date. Optional.
periods_in_a_year_fixed (DaysInYearType | None) – Allows locking the periods-in-a-year to simplify test cases and comparisons. Optional.
self (Self)

Returns:

Information Ratios.

Return type:

Series[float]

OpenFrame.capture_ratio_func(ratio, base_column=-1, months_from_last=None, from_date=None, to_date=None, periods_in_a_year_fixed=None)[source]

Capture Ratio.

The Up (Down) Capture Ratio is calculated by dividing the CAGR of the asset during periods that the benchmark returns are positive (negative) by the CAGR of the benchmark during the same periods. CaptureRatio.BOTH is the Up ratio divided by the Down ratio. Source: ‘Capture Ratios: A Popular Method of Measuring Portfolio Performance in Practice’, Don R. Cox and Delbert C. Goff, Journal of Economics and Finance Education (Vol 2 Winter 2013).

Reference: https://www.economics-finance.org/jefe/volume12-2/11ArticleCox.pdf.

Parameters:

ratio (LiteralCaptureRatio) – The ratio to calculate.
base_column (tuple[str, ValueType] | int) – Column of timeseries that is the denominator in the ratio. Defaults to -1.
months_from_last (int | None) – Number of months offset as positive integer. Overrides use of from_date and to_date. Optional.
from_date (dt.date | None) – Specific from date. Optional.
to_date (dt.date | None) – Specific to date. Optional.
periods_in_a_year_fixed (DaysInYearType | None) – Allows locking the periods-in-a-year to simplify test cases and comparisons. Optional.
self (Self)

Returns:

Capture Ratios.

Return type:

Series[float]

OpenFrame.multi_factor_linear_regression(dependent_column)[source]

Perform a multi-factor linear regression.

This function treats one specified column in the DataFrame as the dependent variable (y) and uses all remaining columns as independent variables (X). It utilizes a scikit-learn LinearRegression model and returns a DataFrame with summary output and an OpenTimeSeries of predicted values.

Parameters:

dependent_column (tuple[str, ValueType]) – A tuple key to select the column in the OpenFrame.tsdf.columns to use as the dependent variable.
self (Self)

Returns:

A DataFrame with the R-squared, the intercept and the regression coefficients
An OpenTimeSeries of predicted values

Return type:

A tuple containing

Raises:

KeyError – If the column tuple is not found in the OpenFrame.tsdf.columns.
ValueError – If not all series are returnseries (ValueType.RTRN).

Rolling Analysis

OpenFrame.rolling_info_ratio(long_column=0, short_column=1, observations=21, periods_in_a_year_fixed=None)[source]

Calculate rolling Information Ratio.

The Information Ratio equals ( fund return less index return ) divided by the Tracking Error. And the Tracking Error is the standard deviation of the difference between the fund and its index returns.

Parameters:

long_column (int) – Column of timeseries that is the numerator in the ratio. Defaults to 0.
short_column (int) – Column of timeseries that is the denominator in the ratio. Defaults to 1.
observations (int) – The length of the rolling window to use is set as number of observations. Defaults to 21.
periods_in_a_year_fixed (Annotated[int, FieldInfo(annotation=NoneType, required=True, metadata=[Strict(strict=True), Ge(ge=1), Le(le=366)])] | None) – Allows locking the periods-in-a-year to simplify test cases and comparisons. Optional.
self (Self)

Returns:

Rolling Information Ratios.

Return type:

OpenFrame.rolling_beta(asset_column=0, market_column=1, observations=21, dlta_degr_freedms=1)[source]

Calculate rolling Market Beta.

Calculates Beta as Co-variance of asset & market divided by Variance of the market.

Reference: https://www.investopedia.com/terms/b/beta.asp.

Parameters:

asset_column (int) – Column of timeseries that is the asset. Defaults to 0.
market_column (int) – Column of timeseries that is the market. Defaults to 1.
observations (int) – The length of the rolling window to use is set as number of observations. Defaults to 21.
dlta_degr_freedms (int) – Variance bias factor taking the value 0 or 1. Defaults to 1.
self (Self)

Returns:

Rolling Betas.

Return type:

OpenFrame.rolling_corr(first_column=0, second_column=1, observations=21)[source]

Calculate rolling Correlation.

Calculates correlation between two series. The period with at least the given number of observations is the first period calculated.

Parameters:

first_column (int) – The position as integer of the first timeseries to compare. Defaults to 0.
second_column (int) – The position as integer of the second timeseries to compare. Defaults to 1.
observations (int) – The length of the rolling window to use is set as number of observations. Defaults to 21.
self (Self)

Returns:

Rolling Correlations.

Return type:

OpenFrame.rolling_return(column=0, observations=21)

Calculate rolling returns.

Parameters:

column (int) – Column position to calculate.
observations (int) – Number of observations in the overlapping window.
self (Self)

Returns:

DataFrame with rolling returns.

Return type:

OpenFrame.rolling_vol(column=0, observations=21, periods_in_a_year_fixed=None, dlta_degr_freedms=1)

Calculate rolling annualized volatilities.

Parameters:

column (int) – Column position to calculate.
observations (int) – Number of observations in the overlapping window.
periods_in_a_year_fixed (DaysInYearType | None) – Lock periods-in-a-year to simplify tests and comparisons.
dlta_degr_freedms (int) – Variance bias factor (0 or 1).
self (Self)

Returns:

DataFrame with rolling annualized volatilities.

Return type:

DataFrame

OpenFrame.rolling_var_down(column=0, level=0.95, observations=252, interpolation='lower')

Calculate rolling annualized downside Value At Risk (VaR).

Parameters:

column (int) – Column position to calculate.
level (float) – Value At Risk level.
observations (int) – Number of observations in the overlapping window.
interpolation (LiteralQuantileInterp) – Interpolation used by DataFrame.quantile.
self (Self)

Returns:

DataFrame with rolling annualized downside VaR.

Return type:

DataFrame

OpenFrame.rolling_cvar_down(column=0, level=0.95, observations=252)

Calculate rolling annualized downside CVaR.

Parameters:

column (int) – Column position to calculate.
level (float) – Conditional Value At Risk level.
observations (int) – Number of observations in the overlapping window.
self (Self)

Returns:

DataFrame with rolling annualized downside CVaR.

Return type:

Correlation and Risk

OpenFrame.correl_matrix

Correlation matrix.

This property returns the correlation matrix of the time series in the frame.

Returns:: Correlation matrix of the time series in the frame.

OpenFrame.ewma_risk(lmbda=0.94, day_chunk=11, dlta_degr_freedms=0, first_column=0, second_column=1, corr_scale=2.0, months_from_last=None, from_date=None, to_date=None, periods_in_a_year_fixed=None)[source]

Exponentially Weighted Moving Average Volatilities and Correlation.

Exponentially Weighted Moving Average (EWMA) for Volatilities and Correlation.

Reference: https://www.investopedia.com/articles/07/ewma.asp.

Parameters:

lmbda (float) – Scaling factor to determine weighting. Defaults to 0.94.
day_chunk (int) – Sampling the data which is assumed to be daily. Defaults to 11.
dlta_degr_freedms (int) – Variance bias factor taking the value 0 or 1. Defaults to 0.
first_column (int) – Column of first timeseries. Defaults to 0.
second_column (int) – Column of second timeseries. Defaults to 1.
corr_scale (float) – Correlation scale factor. Defaults to 2.0.
months_from_last (int | None) – Number of months offset as positive integer. Overrides use of from_date and to_date. Optional.
from_date (dt.date | None) – Specific from date. Optional.
to_date (dt.date | None) – Specific to date. Optional.
periods_in_a_year_fixed (DaysInYearType | None) – Allows locking the periods-in-a-year to simplify test cases and comparisons. Optional.
self (Self)

Returns:

Series volatilities and correlation.

Return type:

DataFrame

Data Manipulation

OpenFrame.align_index_to_local_cdays(countries=None, markets=None, custom_holidays=None, method='nearest')

Align the index of .tsdf with local calendar business days.

Parameters:

countries (CountriesType | None) – Country code(s) (ISO 3166-1 alpha-2).
markets (list[str] | str | None) – Market code(s) supported by exchange_calendars.
custom_holidays (list[str] | str | None) – Missing holidays that should be added.
method (LiteralPandasReindexMethod) – Method for reindexing when aligning to business days.
self (Self)

Returns:

The modified object.

Return type:

Self

OpenFrame.resample(freq='BME')[source]

Resample the timeseries frequency.

Parameters:

freq (Literal['B', 'BME', 'BQE', 'BYE'] | str) – The date offset string that sets the resampled frequency. Defaults to “BME”.
self (Self)

Returns:

An OpenFrame object.

Return type:

OpenFrame.resample_to_business_period_ends(freq='BME', method='nearest')[source]

Resamples timeseries frequency to the business calendar month end dates.

Stubs left in place. Stubs will be aligned to the shortest stub.

Parameters:

freq (Literal['B', 'BME', 'BQE', 'BYE']) – The date offset string that sets the resampled frequency. Defaults to “BME”.
method (Literal['pad', 'ffill', 'backfill', 'bfill', 'nearest'] | None) – Controls the method used to align values across columns. Defaults to nearest.
self (Self)

Returns:

An OpenFrame object.

Return type:

OpenFrame.value_nan_handle(method='fill')

Handle missing values in a value series.

Parameters:

method (LiteralNanMethod) – Method used to handle NaN. Either "fill" (last known) or "drop".
self (Self)

Returns:

The modified object.

Return type:

Self

OpenFrame.return_nan_handle(method='fill')

Handle missing values in a return series.

Parameters:

method (LiteralNanMethod) – Method used to handle NaN. Either "fill" (zero) or "drop".
self (Self)

Returns:

The modified object.

Return type:

Self

Transformations

OpenFrame.to_cumret()[source]

Convert series of returns into cumulative series of values.

Returns:: An OpenFrame object.
Parameters:: self (Self)
Return type:: Self

OpenFrame.value_to_ret()[source]

Convert series of values into series of returns.

Returns:: The returns of the values in the series.
Parameters:: self (Self)
Return type:: Self

OpenFrame.value_to_diff(periods=1)[source]

Convert series of values to series of their period differences.

Parameters:

periods (int) – The number of periods between observations over which difference is calculated. Defaults to 1.
self (Self)

Returns:

An OpenFrame object.

Return type: