Skip to main content


Builds a hierarchical equal risk contribution portfolio


herc [-cd {pearson,spearman,abs_pearson,abs_spearman,distance,mutual_info,tail}] [-cv {hist,ewma1,ewma2,ledoit,oas,shrunk,gl,jlogo,fixed,spectral,shrink}] [-rm RISK-MEASURE] [-as CVAR_SIMULATIONS_LOSSES] [-b CVAR_SIGNIFICANCE] [-bs CVAR_SIMULATIONS_GAINS] [-lk LINKAGE] [-k AMOUNT_CLUSTERS] [-mk MAX_CLUSTERS] [-bi {KN,FD,SC,HGR}] [-at ALPHA_TAIL] [-lo LEAF_ORDER] [-de SMOOTHING_FACTOR_EWMA] [-mt METHOD] [-ct CATEGORIES] [-p PERIOD] [-s START_PERIOD] [-e END_PERIOD] [-lr] [--freq {d,w,m}] [-mn MAX_NAN] [-th THRESHOLD_VALUE] [-r RISK_FREE] [-a SIGNIFICANCE_LEVEL] [-v LONG_ALLOCATION] [--name NAME]


co_dependenceThe codependence or similarity matrix used to build the distance metric and clusters. Possible values are: 'pearson': pearson correlation matrix 'spearman': spearman correlation matrix 'abs_pearson': absolute value of pearson correlation matrix 'abs_spearman': absolute value of spearman correlation matrix 'distance': distance correlation matrix 'mutual_info': mutual information codependence matrix 'tail': tail index codependence matrixpearsonTruepearson, spearman, abs_pearson, abs_spearman, distance, mutual_info, tail
covarianceMethod used to estimate covariance matrix. Possible values are 'hist': historical method 'ewma1': exponential weighted moving average with adjust=True 'ewma2': exponential weighted moving average with adjust=False 'ledoit': Ledoit and Wolf shrinkage method 'oas': oracle shrinkage method 'shrunk': scikit-learn shrunk method 'gl': graphical lasso method 'jlogo': j-logo covariance 'fixed': takes average of eigenvalues above max Marchenko Pastour limit 'spectral': makes zero eigenvalues above max Marchenko Pastour limit 'shrink': Lopez de Prado's book shrinkage methodhistTruehist, ewma1, ewma2, ledoit, oas, shrunk, gl, jlogo, fixed, spectral, shrink
risk_measureRisk measure used to optimize the portfolio. Possible values are: 'MV' : Variance 'MAD' : Mean Absolute Deviation 'GMD' : Gini Mean Difference 'MSV' : Semi Variance (Variance of negative returns) 'FLPM' : First Lower Partial Moment 'SLPM' : Second Lower Partial Moment 'VaR' : Value at Risk 'CVaR' : Conditional Value at Risk 'TG' : Tail Gini 'EVaR' : Entropic Value at Risk 'WR' : Worst Realization 'RG' : Range 'CVRG' : CVaR Range 'TGRG' : Tail Gini Range 'ADD' : Average Drawdown of uncompounded returns 'UCI' : Ulcer Index of uncompounded returns 'DaR' : Drawdown at Risk of uncompounded returns 'CDaR' : Conditional Drawdown at Risk of uncompounded returns 'EDaR' : Entropic Drawdown at Risk of uncompounded returns 'MDD' : Maximum Drawdown of uncompounded returns 'ADD_Rel' : Average Drawdown of compounded returns 'UCI_Rel' : Ulcer Index of compounded returns 'DaR_Rel' : Drawdown at Risk of compounded returns 'CDaR_Rel' : Conditional Drawdown at Risk of compounded returns 'EDaR_Rel' : Entropic Drawdown at Risk of compounded returns 'MDD_Rel' : Maximum Drawdown of compounded returnsMVTrueMV, MAD, GMD, MSV, VaR, CVaR, TG, EVaR, RG, CVRG, TGRG, WR, FLPM, SLPM, MDD, ADD, DaR, CDaR, EDaR, UCI, MDD_Rel, ADD_Rel, DaR_Rel, CDaR_Rel, EDaR_Rel, UCI_Rel
cvar_simulations_lossesNumber of CVaRs used to approximate Tail Gini of losses. The default is 100100TrueNone
cvar_significanceSignificance level of CVaR and Tail Gini of gains. If empty it duplicates alphaNoneTrueNone
cvar_simulations_gainsNumber of CVaRs used to approximate Tail Gini of gains. If empty it duplicates a_sim valueNoneTrueNone
linkageLinkage method of hierarchical clusteringsingleTruesingle, complete, average, weighted, centroid, median, ward, dbht
amount_clustersNumber of clusters specified in advanceNoneTrueNone
max_clustersMax number of clusters used by the two difference gap statistic to find the optimal number of clusters. If k is empty this value is used10TrueNone
amount_binsNumber of bins used to calculate the variation of informationKNTrueKN, FD, SC, HGR
alpha_tailSignificance level for lower tail dependence index, only used when when codependence value is 'tail'0.05TrueNone
leaf_orderIndicates if the cluster are ordered so that the distance between successive leaves is minimalTrueTrueNone
smoothing_factor_ewmaSmoothing factor for ewma estimators0.94TrueNone
nan_fill_methodMethod used to fill nan values in time series, by default time. Possible values are: 'linear': linear interpolation 'time': linear interpolation based on time index 'nearest': use nearest value to replace nan values 'zero': spline of zeroth order 'slinear': spline of first order 'quadratic': spline of second order 'cubic': spline of third order 'barycentric': builds a polynomial that pass for all pointstimeTruelinear, time, nearest, zero, slinear, quadratic, cubic, barycentric
categoriesShow selected categoriesASSET_CLASS, COUNTRY, SECTOR, INDUSTRYTrueNone
historic_periodPeriod to get yfinance data from. Possible frequency strings are: 'd': means days, for example '252d' means 252 days 'w': means weeks, for example '52w' means 52 weeks 'mo': means months, for example '12mo' means 12 months 'y': means years, for example '1y' means 1 year 'ytd': downloads data from beginning of year to today 'max': downloads all data available for each asset3yTrue1d, 5d, 1mo, 3mo, 6mo, 1y, 2y, 5y, 10y, ytd, max
start_periodStart date to get yfinance data from. Must be in 'YYYY-MM-DD' formatTrueNone
end_periodEnd date to get yfinance data from. Must be in 'YYYY-MM-DD' formatTrueNone
log_returnsIf use logarithmic or arithmetic returns to calculate returnsFalseTrueNone
return_frequencyFrequency used to calculate returns. Possible values are: 'd': for daily returns 'w': for weekly returns 'm': for monthly returnsdTrued, w, m
max_nanMax percentage of nan values accepted per asset to be considered in the optimization process0.05TrueNone
threshold_valueValue used to replace outliers that are higher to threshold in absolute value0.3TrueNone
risk_freeRisk-free rate of borrowing/lending. The period of the risk-free rate must be annual0.02924TrueNone
significance_levelSignificance level of CVaR, EVaR, CDaR and EDaR0.05TrueNone
long_allocationAmount to allocate to portfolio1TrueNone
nameSave portfolio with personalized or default nameHERC_0TrueNone


2022 Apr 05, 14:40 (๐Ÿฆ‹) /portfolio/po/ $ herc

[3 Years] Hierarchical equal risk contribution portfolio using pearson
codependence,single linkage and volatility as risk measure

โ”ƒ โ”ƒ Value โ”ƒ
โ”‚ AAPL โ”‚ 10.78 % โ”‚
โ”‚ AMZN โ”‚ 12.64 % โ”‚
โ”‚ BA โ”‚ 9.12 % โ”‚
โ”‚ FB โ”‚ 8.08 % โ”‚
โ”‚ MSFT โ”‚ 12.86 % โ”‚
โ”‚ T โ”‚ 43.83 % โ”‚
โ”‚ TSLA โ”‚ 2.66 % โ”‚

Annual (by 252) expected return: 18.09%
Annual (by โˆš252) volatility: 24.19%
Sharpe ratio: 0.7401