def factors_stats(self, method_mu="hist", method_cov="hist", **kwargs):
r"""
Calculate the inputs that will be use by the optimization method when
we select the input model='FM'.
Parameters
----------
**kwargs : dict
All aditional parameters of risk_factors function.
See Also
--------
riskfolio.ParamsEstimation.forward_regression
riskfolio.ParamsEstimation.backward_regression
riskfolio.ParamsEstimation.loadings_matrix
riskfolio.ParamsEstimation.risk_factors
"""
X = self.factors
Y = self.returns
mu, cov, returns, nav = pe.risk_factors(
X, Y, method_mu=method_mu, method_cov=method_cov, **kwargs
)
self.mu_fm = mu
self.cov_fm = cov
self.returns_fm = returns
self.nav_fm = nav
value = af.is_pos_def(self.cov_fm, threshold=1e-8)
if value == False:
print("You must convert self.cov_fm to a positive definite matrix")
def _hierarchical_clustering_herc(self,
linkage="ward",
max_k=10,
leaf_order=True):
# hierarchcial clustering
dist = np.sqrt((1 - self.corr).round(8) / 2)
dist = pd.DataFrame(dist,
columns=self.corr.columns,
index=self.corr.index)
p_dist = squareform(dist, checks=False)
clustering = hr.linkage(p_dist,
method=linkage,
optimal_ordering=leaf_order)
# optimal number of clusters
k = af.two_diff_gap_stat(self.corr, dist, clustering, max_k)
return clustering, k
def assets_stats(self, method_mu="hist", method_cov="hist", **kwargs):
r"""
Calculate the inputs that will be use by the optimization method when
we select the input model='Classic'.
Parameters
----------
**kwargs : dict
All aditional parameters of mean_vector and covar_matrix functions.
See Also
--------
riskfolio.ParamsEstimation.mean_vector
riskfolio.ParamsEstimation.covar_matrix
"""
self.mu = pe.mean_vector(self.returns, method=method_mu, **kwargs)
self.cov = pe.covar_matrix(self.returns, method=method_cov, **kwargs)
value = af.is_pos_def(self.cov, threshold=1e-8)
if value == False:
print("You must convert self.cov to a positive definite matrix")
def optimization(
self,
model="HRP",
correlation="pearson",
covariance="hist",
rm="MV",
rf=0,
linkage="single",
k=None,
max_k=10,
leaf_order=True,
d=0.94,
):
r"""
This method calculates the optimal portfolio according to the
optimization model selected by the user.
Parameters
----------
model : str can be {'HRP', 'HERC' or 'HERC2'}
The hierarchical cluster portfolio model used for optimize the
portfolio. The default is 'HRP'. Posible values are:
- 'HRP': Hierarchical Risk Parity.
- 'HERC': Hierarchical Equal Risk Contribution.
- 'HERC2': HERC but splitting weights equally within clusters.
correlation : str can be {'pearson', 'spearman' or 'distance'}.
The correlation matrix used for create the clusters.
The default is 'pearson'. Posible values are:
- 'pearson': pearson correlation matrix.
- 'spearman': spearman correlation matrix.
- 'abs_pearson': absolute value pearson correlation matrix.
- 'abs_spearman': absolute value spearman correlation matrix.
- 'distance': distance correlation matrix.
covariance : str, can be {'hist', 'ewma1', 'ewma2', 'ledoit', 'oas' or 'shrunk'}
The method used to estimate the covariance matrix:
The default is 'hist'.
- 'hist': use historical estimates.
- 'ewma1'': use ewma with adjust=True, see `EWM <https://pandas.pydata.org/pandas-docs/stable/user_guide/computation.html#exponentially-weighted-windows>`_ for more details.
- 'ewma2': use ewma with adjust=False, see `EWM <https://pandas.pydata.org/pandas-docs/stable/user_guide/computation.html#exponentially-weighted-windows>`_ for more details.
- 'ledoit': use the Ledoit and Wolf Shrinkage method.
- 'oas': use the Oracle Approximation Shrinkage method.
- 'shrunk': use the basic Shrunk Covariance method.
rm : str, optional
The risk measure used to optimze the portfolio.
The default is 'MV'. Posible values are:
- 'equal': Equally weighted.
- 'vol': Standard Deviation.
- 'MV': Variance.
- 'MAD': Mean Absolute Deviation.
- 'MSV': Semi Standard Deviation.
- 'FLPM': First Lower Partial Moment (Omega Ratio).
- 'SLPM': Second Lower Partial Moment (Sortino Ratio).
- 'VaR': Value at Risk.
- 'CVaR': Conditional Value at Risk.
- 'EVaR': Entropic Value at Risk.
- 'WR': Worst Realization (Minimax)
- 'MDD': Maximum Drawdown of uncompounded cumulative returns (Calmar Ratio).
- 'ADD': Average Drawdown of uncompounded cumulative returns.
- 'DaR': Drawdown at Risk of uncompounded cumulative returns.
- 'CDaR': Conditional Drawdown at Risk of uncompounded cumulative returns.
- 'EDaR': Entropic Drawdown at Risk of uncompounded cumulative returns.
- 'UCI': Ulcer Index of uncompounded cumulative returns.
- 'MDD_Rel': Maximum Drawdown of compounded cumulative returns (Calmar Ratio).
- 'ADD_Rel': Average Drawdown of compounded cumulative returns.
- 'DaR_Rel': Drawdown at Risk of compounded cumulative returns.
- 'CDaR_Rel': Conditional Drawdown at Risk of compounded cumulative returns.
- 'EDaR_Rel': Entropic Drawdown at Risk of compounded cumulative returns.
- 'UCI_Rel': Ulcer Index of compounded cumulative returns.
rf : float, optional
Risk free rate, must be in the same period of assets returns.
The default is 0.
linkage : string, optional
Linkage method of hierarchical clustering, see `linkage <https://docs.scipy.org/doc/scipy/reference/generated/scipy.cluster.hierarchy.linkage.html?highlight=linkage#scipy.cluster.hierarchy.linkage>`_ for more details.
The default is 'single'. Posible values are:
- 'single'.
- 'complete'.
- 'average'.
- 'weighted'.
- 'centroid'.
- 'median'.
- 'ward'.
k : int, optional
Number of clusters. This value is took instead of the optimal number
of clusters calculated with the two difference gap statistic.
The default is None.
max_k : int, optional
Max number of clusters used by the two difference gap statistic
to find the optimal number of clusters. The default is 10.
leaf_order : bool, optional
Indicates if the cluster are ordered so that the distance between
successive leaves is minimal. The default is True.
d : scalar
The smoothing factor of ewma methods.
The default is 0.94.
Returns
-------
w : DataFrame
The weights of optimal portfolio.
"""
# Correlation matrix from covariance matrix
self.cov = pe.covar_matrix(self.returns, method=covariance, d=0.94)
if correlation in {"pearson", "spearman"}:
self.corr = self.returns.corr(method=correlation).astype(float)
if correlation in {"abs_pearson", "abs_spearman"}:
self.corr = np.abs(
self.returns.corr(method=correlation[4:])).astype(float)
elif correlation == "distance":
self.corr = af.dcorr_matrix(self.returns).astype(float)
# Step-1: Tree clustering
if model == "HRP":
self.clusters = self._hierarchical_clustering_hrp(
linkage, leaf_order=leaf_order)
elif model in ["HERC", "HERC2"]:
self.clusters, self.k = self._hierarchical_clustering_herc(
linkage, max_k, leaf_order=leaf_order)
if k is not None:
self.k = int(k)
# Step-2: Seriation (Quasi-Diagnalization)
self.sort_order = self._seriation(self.clusters)
asset_order = self.assetslist
asset_order[:] = [self.assetslist[i] for i in self.sort_order]
self.asset_order = asset_order
self.corr_sorted = self.corr.reindex(index=self.asset_order,
columns=self.asset_order)
# Step-3: Recursive bisection
if model == "HRP":
weights = self._recursive_bisection(self.sort_order, rm=rm, rf=rf)
elif model in ["HERC", "HERC2"]:
weights = self._hierarchical_recursive_bisection(self.clusters,
rm=rm,
rf=rf,
linkage=linkage,
model=model)
weights = weights.loc[self.assetslist].to_frame()
weights.columns = ["weights"]
return weights
def assets_clusters(returns,
correlation="pearson",
linkage="ward",
k=None,
max_k=10,
leaf_order=True):
r"""
Create asset classes based on hierarchical clustering.
Parameters
----------
returns : DataFrame
Assets returns.
correlation : str can be {'pearson', 'spearman' or 'distance'}.
The correlation matrix used for create the clusters.
The default is 'pearson'. Posible values are:
- 'pearson': pearson correlation matrix.
- 'spearman': spearman correlation matrix.
- 'abs_pearson': absolute value pearson correlation matrix.
- 'abs_spearman': absolute value spearman correlation matrix.
- 'distance': distance correlation matrix.
linkage : string, optional
Linkage method of hierarchical clustering, see `linkage <https://docs.scipy.org/doc/scipy/reference/generated/scipy.cluster.hierarchy.linkage.html?highlight=linkage#scipy.cluster.hierarchy.linkage>`_ for more details.
The default is 'single'. Posible values are:
- 'single'.
- 'complete'.
- 'average'.
- 'weighted'.
- 'centroid'.
- 'median'.
- 'ward'.
k : int, optional
Number of clusters. This value is took instead of the optimal number
of clusters calculated with the two difference gap statistic.
The default is None.
max_k : int, optional
Max number of clusters used by the two difference gap statistic
to find the optimal number of clusters. The default is 10.
leaf_order : bool, optional
Indicates if the cluster are ordered so that the distance between
successive leaves is minimal. The default is True.
Returns
-------
clusters : DataFrame
A dataframe with asset classes based on hierarchical clustering.
Raises
------
ValueError when the value cannot be calculated.
Examples
--------
::
clusters = cf.assets_clusters(returns, correlation='pearson', linkage='ward', k=None, max_k=10, leaf_order=True)
The clusters dataframe looks like this:
.. image:: images/clusters_df.png
"""
if not isinstance(returns, pd.DataFrame):
raise ValueError("returns must be a DataFrame")
# Correlation matrix from covariance matrix
if correlation in {"pearson", "spearman"}:
corr = returns.corr(method=correlation)
if correlation in {"abs_pearson", "abs_spearman"}:
corr = np.abs(returns.corr(method=correlation[4:]))
elif correlation == "distance":
corr = af.dcorr_matrix(returns)
# hierarchcial clustering
dist = np.sqrt((1 - corr).round(8) / 2)
dist = pd.DataFrame(dist, columns=corr.columns, index=corr.index)
p_dist = squareform(dist, checks=False)
clustering = hr.linkage(p_dist,
method=linkage,
optimal_ordering=leaf_order)
if k is None:
# optimal number of clusters
k = af.two_diff_gap_stat(corr, dist, clustering, max_k)
clusters_inds = hr.fcluster(clustering, k, criterion="maxclust")
clusters = {"Assets": [], "Clusters": []}
for i, v in enumerate(clusters_inds):
clusters["Assets"].append(corr.columns.tolist()[i])
clusters["Clusters"].append("Cluster " + str(v))
clusters = pd.DataFrame(clusters)
clusters = clusters.sort_values(by=["Assets"])
return clusters
def optimization(
self, model="Classic", rm="MV", obj="Sharpe", rf=0, l=2, hist=True
):
r"""
This method that calculates the optimum portfolio according to the
optimization model selected by the user. The general problem that
solves is:
.. math::
\begin{align}
&\underset{x}{\text{optimize}} & & F(w)\\
&\text{s. t.} & & Aw \geq B\\
& & & \phi_{i}(w) \leq c_{i}\\
\end{align}
Where:
:math:`F(w)` is the objective function.
:math:`Aw \geq B` is a set of linear constraints.
:math:`\phi_{i}(w) \leq c_{i}` are constraints on maximum values of
several risk measures.
Parameters
----------
model : str can be 'Classic', 'BL' or 'FM'
The model used for optimize the portfolio.
The default is 'Classic'. Posible values are:
- 'Classic': use estimates of expected return vector and covariance matrix that depends on historical data.
- 'BL': use estimates of expected return vector and covariance matrix based on the Black Litterman model.
- 'FM': use estimates of expected return vector and covariance matrix based on a Risk Factor model specified by the user.
rm : str, optional
The risk measure used to optimze the portfolio.
The default is 'MV'. Posible values are:
- 'MV': Standard Deviation.
- 'MAD': Mean Absolute Deviation.
- 'MSV': Semi Standard Deviation.
- 'FLPM': First Lower Partial Moment (Omega Ratio).
- 'SLPM': Second Lower Partial Moment (Sortino Ratio).
- 'CVaR': Conditional Value at Risk.
- 'WR': Worst Realization (Minimax)
- 'MDD': Maximum Drawdown of uncompounded returns (Calmar Ratio).
- 'ADD': Average Drawdown of uncompounded returns.
- 'CDaR': Conditional Drawdown at Risk of uncompounded returns.
obj : str can be {'MinRisk', 'Utility', 'Sharpe' or 'MaxRet'.
Objective function of the optimization model.
The default is 'Sharpe'. Posible values are:
- 'MinRisk': Minimize the selected risk measure.
- 'Utility': Maximize the Utility function :math:`mu w - l \phi_{i}(w)`.
- 'Sharpe': Maximize the risk adjusted return ratio based on the selected risk measure.
- 'MaxRet': Maximize the expected return of the portfolio.
rf : float, optional
Risk free rate, must be in the same period of assets returns.
The default is 0.
l : scalar, optional
Risk aversion factor of the 'Utility' objective function.
The default is 2.
hist : bool, optional
Indicate if uses historical or factor estimation of returns to
calculate risk measures that depends on scenarios (All except
'MV' risk measure). The default is True.
Returns
-------
w : DataFrame
The weights of optimum portfolio.
"""
# General model Variables
mu = None
sigma = None
returns = None
if model == "Classic":
mu = np.matrix(self.mu)
sigma = np.matrix(self.cov)
returns = np.matrix(self.returns)
nav = np.matrix(self.nav)
elif model == "FM":
mu = np.matrix(self.mu_fm)
if hist == False:
sigma = np.matrix(self.cov_fm)
returns = np.matrix(self.returns_fm)
nav = np.matrix(self.nav_fm)
elif hist == True:
sigma = np.matrix(self.cov)
returns = np.matrix(self.returns)
nav = np.matrix(self.nav)
elif model == "BL":
mu = np.matrix(self.mu_bl)
if hist == False:
sigma = np.matrix(self.cov_bl)
elif hist == True:
sigma = np.matrix(self.cov)
returns = np.matrix(self.returns)
nav = np.matrix(self.nav)
elif model == "BL_FM":
mu = np.matrix(self.mu_bl_fm)
if hist == False:
sigma = np.matrix(self.cov_bl_fm)
returns = np.matrix(self.returns_fm)
nav = np.matrix(self.nav_fm)
elif hist == True:
sigma = np.matrix(self.cov)
returns = np.matrix(self.returns)
nav = np.matrix(self.nav)
# General Model Variables
returns = np.matrix(returns)
w = cv.Variable((mu.shape[1], 1))
k = cv.Variable((1, 1))
rf0 = cv.Parameter(nonneg=True)
rf0.value = rf
n = cv.Parameter(nonneg=True)
n.value = returns.shape[0]
ret = mu * w
# MV Model Variables
risk1 = cv.quad_form(w, sigma)
returns_1 = af.cov_returns(sigma) * 1000
n1 = cv.Parameter(nonneg=True)
n1.value = returns_1.shape[0]
risk1_1 = cv.norm(returns_1 * w, "fro") / cv.sqrt(n1 - 1)
# MAD Model Variables
madmodel = False
Y = cv.Variable((returns.shape[0], 1))
u = np.matrix(np.ones((returns.shape[0], 1)) * mu)
a = returns - u
risk2 = cv.sum(Y) / n
# madconstraints=[a*w >= -Y, a*w <= Y, Y >= 0]
madconstraints = [a * w <= Y, Y >= 0]
# Semi Variance Model Variables
risk3 = cv.norm(Y, "fro") / cv.sqrt(n - 1)
# CVaR Model Variables
alpha1 = self.alpha
VaR = cv.Variable(1)
alpha = cv.Parameter(nonneg=True)
alpha.value = alpha1
X = returns * w
Z = cv.Variable((returns.shape[0], 1))
risk4 = VaR + 1 / (alpha * n) * cv.sum(Z)
cvarconstraints = [Z >= 0, Z >= -X - VaR]
# Worst Realization (Minimax) Model Variables
M = cv.Variable(1)
risk5 = M
wrconstraints = [-X <= M]
# Lower Partial Moment Variables
lpmmodel = False
lpm = cv.Variable((returns.shape[0], 1))
lpmconstraints = [lpm >= 0]
if obj == "Sharpe":
lpmconstraints += [lpm >= rf0 * k - X]
else:
lpmconstraints += [lpm >= rf0 - X]
# First Lower Partial Moment (Omega) Model Variables
risk6 = cv.sum(lpm) / n
# Second Lower Partial Moment (Sortino) Model Variables
risk7 = cv.norm(lpm, "fro") / cv.sqrt(n - 1)
# Drawdown Model Variables
drawdown = False
if obj == "Sharpe":
X1 = k + nav * w
else:
X1 = 1 + nav * w
U = cv.Variable((nav.shape[0] + 1, 1))
ddconstraints = [U[1:] >= X1, U[1:] >= U[:-1]]
if obj == "Sharpe":
ddconstraints += [U[1:] >= k, U[0] == k]
else:
ddconstraints += [U[1:] >= 1, U[0] == 1]
# Maximum Drawdown Model Variables
MDD = cv.Variable(1)
risk8 = MDD
mddconstraints = [MDD >= U[1:] - X1]
# Average Drawdown Model Variables
risk9 = 1 / n * cv.sum(U[1:] - X1)
# Conditional Drawdown Model Variables
CDaR = cv.Variable(1)
Zd = cv.Variable((nav.shape[0], 1))
risk10 = CDaR + 1 / (alpha * n) * cv.sum(Zd)
cdarconstraints = [Zd >= U[1:] - X1 - CDaR, Zd >= 0]
# Tracking Error Model Variables
c = self.benchweights
if self.kindbench == True:
bench = np.matrix(returns) * c
else:
bench = self.benchindex
if obj == "Sharpe":
TE = cv.norm(returns * w - bench * k, "fro") / cv.sqrt(n - 1)
else:
TE = cv.norm(returns * w - bench, "fro") / cv.sqrt(n - 1)
# Problem aditional linear constraints
if obj == "Sharpe":
constraints = [
cv.sum(w) == self.upperlng * k,
k >= 0,
mu * w - rf0 * k == 1,
]
if self.sht == False:
constraints += [w <= self.upperlng * k, w * 1000 >= 0]
elif self.sht == True:
constraints += [
w <= self.upperlng * k,
w >= -self.uppersht * k,
cv.sum(cv.neg(w)) <= self.uppersht * k,
]
else:
constraints = [cv.sum(w) == self.upperlng]
if self.sht == False:
constraints += [w <= self.upperlng, w * 1000 >= 0]
elif self.sht == True:
constraints += [
w <= self.upperlng,
w >= -self.uppersht,
cv.sum(cv.neg(w)) <= self.uppersht,
]
if self.ainequality is not None and self.binequality is not None:
A = np.matrix(self.ainequality)
B = np.matrix(self.binequality)
if obj == "Sharpe":
constraints += [A * w - B * k >= 0]
else:
constraints += [A * w - B >= 0]
# Turnover Constraints
if obj == "Sharpe":
if self.allowTO == True:
constraints += [cv.abs(w - c * k) * 1000 <= self.turnover * k * 1000]
else:
if self.allowTO == True:
constraints += [cv.abs(w - c) * 1000 <= self.turnover * 1000]
# Tracking error Constraints
if obj == "Sharpe":
if self.allowTE == True:
constraints += [TE <= self.TE * k]
else:
if self.allowTE == True:
constraints += [TE <= self.TE]
# Problem risk Constraints
if self.upperdev is not None:
if obj == "Sharpe":
constraints += [risk1_1 <= self.upperdev * k]
else:
constraints += [risk1 <= self.upperdev ** 2]
if self.uppermad is not None:
if obj == "Sharpe":
constraints += [risk2 <= self.uppermad * k / 2]
else:
constraints += [risk2 <= self.uppermad / 2]
madmodel = True
if self.uppersdev is not None:
if obj == "Sharpe":
constraints += [risk3 <= self.uppersdev * k]
else:
constraints += [risk3 <= self.uppersdev]
madmodel = True
if self.upperCVaR is not None:
if obj == "Sharpe":
constraints += [risk4 <= self.upperCVaR * k]
else:
constraints += [risk4 <= self.upperCVaR]
constraints += cvarconstraints
if self.upperwr is not None:
if obj == "Sharpe":
constraints += [-X <= self.upperwr * k]
else:
constraints += [-X <= self.upperwr]
constraints += wrconstraints
if self.upperflpm is not None:
if obj == "Sharpe":
constraints += [risk6 <= self.upperflpm * k]
else:
constraints += [risk6 <= self.upperflpm]
lpmmodel = True
if self.upperslpm is not None:
if obj == "Sharpe":
constraints += [risk7 <= self.upperslpm * k]
else:
constraints += [risk7 <= self.upperslpm]
lpmmodel = True
if self.uppermdd is not None:
if obj == "Sharpe":
constraints += [U[1:] - X1 <= self.uppermdd * k]
else:
constraints += [U[1:] - X1 <= self.uppermdd]
constraints += mddconstraints
drawdown = True
if self.upperadd is not None:
if obj == "Sharpe":
constraints += [risk9 <= self.upperadd * k]
else:
constraints += [risk9 <= self.upperadd]
drawdown = True
if self.upperCDaR is not None:
if obj == "Sharpe":
constraints += [risk10 <= self.upperCDaR * k]
else:
constraints += [risk10 <= self.upperCDaR]
constraints += cdarconstraints
drawdown = True
# Defining risk function
if rm == "MV":
if model != "Classic":
risk = risk1_1
elif model == "Classic":
risk = risk1
elif rm == "MAD":
risk = risk2
madmodel = True
elif rm == "MSV":
risk = risk3
madmodel = True
elif rm == "CVaR":
risk = risk4
if self.upperCVaR is None:
constraints += cvarconstraints
elif rm == "WR":
risk = risk5
if self.upperwr is None:
constraints += wrconstraints
elif rm == "FLPM":
risk = risk6
lpmmodel = True
elif rm == "SLPM":
risk = risk7
lpmmodel = True
elif rm == "MDD":
risk = risk8
drawdown = True
if self.uppermdd is None:
constraints += mddconstraints
elif rm == "ADD":
risk = risk9
drawdown = True
elif rm == "CDaR":
risk = risk10
drawdown = True
if self.upperCDaR is None:
constraints += cdarconstraints
if madmodel == True:
constraints += madconstraints
if lpmmodel == True:
constraints += lpmconstraints
if drawdown == True:
constraints += ddconstraints
# Frontier Variables
portafolio = {}
for i in self.assetslist:
portafolio.update({i: []})
# Optimization Process
# Defining solvers
solvers = [cv.ECOS, cv.SCS, cv.OSQP, cv.CVXOPT, cv.GLPK]
# Defining objective function
if obj == "Sharpe":
if rm != "Classic":
objective = cv.Minimize(risk)
elif rm == "Classic":
objective = cv.Minimize(risk * 1000)
elif obj == "MinRisk":
objective = cv.Minimize(risk)
elif obj == "Utility":
objective = cv.Maximize(ret - l * risk)
elif obj == "MaxRet":
objective = cv.Maximize(ret)
try:
prob = cv.Problem(objective, constraints)
for solver in solvers:
try:
prob.solve(
solver=solver, parallel=True, max_iters=2000, abstol=1e-10
)
except:
pass
if w.value is not None:
break
if obj == "Sharpe":
weights = np.matrix(w.value / k.value).T
else:
weights = np.matrix(w.value).T
if self.sht == False:
weights = np.abs(weights) / np.sum(np.abs(weights))
for j in self.assetslist:
portafolio[j].append(weights[0, self.assetslist.index(j)])
except:
pass
optimum = pd.DataFrame(portafolio, index=["weights"], dtype=np.float64).T
return optimum
def blacklitterman_stats(
self,
P,
Q,
rf=0,
w=None,
delta=None,
eq=True,
method_mu="hist",
method_cov="hist",
**kwargs
):
r"""
Calculate the inputs that will be use by the optimization method when
we select the input model='BL'.
Parameters
----------
P : DataFrame of shape (n_views, n_assets)
Analyst's views matrix, can be relative or absolute.
Q: DataFrame of shape (n_views, 1)
Expected returns of analyst's views.
delta: float
Risk aversion factor. The default value is 1.
rf: scalar, optional
Risk free rate. The default is 0.
w : DataFrame of shape (n_assets, 1)
Weights matrix, where n_assets is the number of assets.
The default is None.
eq: bool, optional
Indicates if use equilibrum or historical excess returns.
The default is True.
**kwargs : dict
Other variables related to the mean and covariance estimation.
See Also
--------
riskfolio.ParamsEstimation.black_litterman
"""
X = self.returns
if w is None:
w = self.benchweights
if delta is None:
a = np.matrix(self.mu) * np.matrix(w)
delta = (a - rf) / (np.matrix(w).T * np.matrix(self.cov) * np.matrix(w))
delta = delta.item()
mu, cov, w = pe.black_litterman(
X=X,
w=w,
P=P,
Q=Q,
delta=delta,
rf=rf,
eq=eq,
method_mu=method_mu,
method_cov=method_cov,
**kwargs
)
self.mu_bl = mu
self.cov_bl = cov
value = af.is_pos_def(self.cov_bl, threshold=1e-8)
if value == False:
print("You must convert self.cov_bl to a positive definite matrix")
def optimization(
self,
model="HRP",
correlation="pearson",
rm="MV",
rf=0,
linkage="single",
k=None,
max_k=10,
leaf_order=True,
):
r"""
This method calculates the optimal portfolio according to the
optimization model selected by the user.
Parameters
----------
model : str can be {'HRP' or 'HERC'}
The hierarchical cluster portfolio model used for optimize the
portfolio. The default is 'HRP'. Posible values are:
- 'HRP': Hierarchical Risk Parity.
- 'HERC': Hierarchical Equal Risk Contribution.
correlation : str can be {'pearson', 'spearman' or 'distance'}.
The correlation matrix used for create the clusters.
The default is 'pearson'. Posible values are:
- 'pearson': pearson correlation matrix.
- 'spearman': spearman correlation matrix.
- 'abs_pearson': absolute value pearson correlation matrix.
- 'abs_spearman': absolute value spearman correlation matrix.
- 'distance': distance correlation matrix.
rm : str, optional
The risk measure used to optimze the portfolio.
The default is 'MV'. Posible values are:
- 'vol': Standard Deviation.
- 'MV': Variance.
- 'MAD': Mean Absolute Deviation.
- 'MSV': Semi Standard Deviation.
- 'FLPM': First Lower Partial Moment (Omega Ratio).
- 'SLPM': Second Lower Partial Moment (Sortino Ratio).
- 'VaR': Value at Risk.
- 'CVaR': Conditional Value at Risk.
- 'EVaR': Entropic Value at Risk.
- 'WR': Worst Realization (Minimax)
- 'MDD': Maximum Drawdown of uncompounded cumulative returns (Calmar Ratio).
- 'ADD': Average Drawdown of uncompounded cumulative returns.
- 'DaR': Drawdown at Risk of uncompounded cumulative returns.
- 'CDaR': Conditional Drawdown at Risk of uncompounded cumulative returns.
- 'EDaR': Entropic Drawdown at Risk of uncompounded cumulative returns.
- 'UCI': Ulcer Index of uncompounded cumulative returns.
- 'MDD_Rel': Maximum Drawdown of compounded cumulative returns (Calmar Ratio).
- 'ADD_Rel': Average Drawdown of compounded cumulative returns.
- 'DaR_Rel': Drawdown at Risk of compounded cumulative returns.
- 'CDaR_Rel': Conditional Drawdown at Risk of compounded cumulative returns.
- 'EDaR_Rel': Entropic Drawdown at Risk of compounded cumulative returns.
- 'UCI_Rel': Ulcer Index of compounded cumulative returns.
rf : float, optional
Risk free rate, must be in the same period of assets returns.
The default is 0.
linkage : string, optional
Linkage method of hierarchical clustering, see `linkage <https://docs.scipy.org/doc/scipy/reference/generated/scipy.cluster.hierarchy.linkage.html?highlight=linkage#scipy.cluster.hierarchy.linkage>`_ for more details.
The default is 'single'. Posible values are:
- 'single'.
- 'complete'.
- 'average'.
- 'weighted'.
- 'centroid'.
- 'median'.
- 'ward'.
k : int, optional
Number of clusters. This value is took instead of the optimal number
of clusters calculated with the two difference gap statistic.
The default is None.
max_k : int, optional
Max number of clusters used by the two difference gap statistic
to find the optimal number of clusters. The default is 10.
Returns
-------
w : DataFrame
The weights of optimal portfolio.
"""
# Correlation matrix from covariance matrix
self.cov = self.returns.cov()
if correlation in {"pearson", "spearman"}:
self.corr = self.returns.corr(method=correlation)
if correlation in {"abs_pearson", "abs_spearman"}:
self.corr = np.abs(self.returns.corr(method=correlation[4:]))
elif correlation == "distance":
self.corr = af.dcorr_matrix(self.returns)
# Step-1: Tree clustering
if model == "HRP":
self.clusters = self._hierarchical_clustering_hrp(
linkage, leaf_order=leaf_order)
elif model == "HERC":
self.clusters, self.k = self._hierarchical_clustering_herc(
linkage, max_k, leaf_order=leaf_order)
if k is not None:
self.k = int(k)
# Step-2: Seriation (Quasi-Diagnalization)
self.sort_order = self._seriation(self.clusters)
asset_order = self.assetslist
asset_order[:] = [self.assetslist[i] for i in self.sort_order]
self.asset_order = asset_order
self.corr_sorted = self.corr.reindex(index=self.asset_order,
columns=self.asset_order)
# Step-3: Recursive bisection
if model == "HRP":
weights = self._recursive_bisection(self.sort_order, rm=rm, rf=rf)
elif model == "HERC":
weights = self._hierarchical_recursive_bisection(self.clusters,
rm=rm,
rf=rf,
linkage=linkage)
weights = weights.loc[self.assetslist].to_frame()
weights.columns = ["weights"]
return weights
def bootstrapping(X, kind='stationary', q=0.05, n_sim=3000, window=3, seed=0):
r"""
Estimates the uncertainty sets of mean and covariance matrix through the selected
bootstrapping method.
Parameters
----------
X : DataFrame of shape (n_samples, n_features)
Features matrix, where n_samples is the number of samples and
n_features is the number of features.
kind : str
The bootstrapping method. The default value is 'stationary'. Posible values are:
- 'stationary': stationary bootstrapping method, see `StationaryBootstrap <https://bashtage.github.io/arch/bootstrap/generated/arch.bootstrap.StationaryBootstrap.html#arch.bootstrap.StationaryBootstrap>`_ for more details.
- 'circular': circular bootstrapping method, see `CircularBlockBootstrap <https://bashtage.github.io/arch/bootstrap/generated/arch.bootstrap.CircularBlockBootstrap.html#arch.bootstrap.CircularBlockBootstrap>`_ for more details.
- 'moving': moving bootstrapping method, see `MovingBlockBootstrap <https://bashtage.github.io/arch/bootstrap/generated/arch.bootstrap.MovingBlockBootstrap.html#arch.bootstrap.MovingBlockBootstrap>`_ for more details.
q : scalar
Significance level of the selected bootstrapping method.
The default is 0.05.
n_sim : scalar
Number of simulations of the bootstrapping method.
The default is 3000.
window:
Block size of the bootstrapping method. Must be greather than 1
and lower than the n_samples - n_features + 1
The default is 3.
seed:
Seed used to generate random numbers for bootstrapping method.
The default is 0.
Returns
-------
mu_l : DataFrame
The q/2 percentile of mean vector obtained through the selected bootstrapping method.
mu_u : DataFrame
The 1-q/2 percentile of mean vector obtained through the selected bootstrapping method.
cov_l : DataFrame
The q/2 percentile of covariance matrix obtained through the selected bootstrapping method.
cov_u : DataFrame
The 1-q/2 percentile of covariance matrix obtained through the selected bootstrapping method.
cov_mu : DataFrame
The covariance matrix of estimation errors of mean vector obtained through the selected bootstrapping method.
We take the diagonal of this matrix following :cite:`b-fabozzi2007robust`.
Raises
------
ValueError
When the value cannot be calculated.
"""
if not isinstance(X, pd.DataFrame):
raise ValueError("X must be a DataFrame")
if window >= X.shape[0] - window + 1:
raise ValueError("block must be lower than n_samples - window + 1")
elif window <= 1:
raise ValueError("block must be greather than 1")
rs = np.random.RandomState(seed)
cols = X.columns.tolist()
m = len(cols)
mus = np.zeros((n_sim, 1, m))
covs = np.zeros((n_sim, m, m))
if kind == 'stationary':
gen = bs.StationaryBootstrap(window, X, random_state=rs)
elif kind == 'circular':
gen = bs.CircularBlockBootstrap(window, X, random_state=rs)
elif kind == 'moving':
gen = bs.MovingBlockBootstrap(window, X, random_state=rs)
else:
raise ValueError(
"kind only can be 'stationary', 'circular' or 'moving'")
i = 0
for data in gen.bootstrap(n_sim):
A = data[0][0]
mus[i] = A.mean().to_numpy().reshape(1, m)
covs[i] = A.cov().to_numpy()
i += 1
mu_l = np.percentile(mus, q / 2 * 100, axis=0, keepdims=True).reshape(1, m)
mu_u = np.percentile(mus, 100 - q / 2 * 100, axis=0,
keepdims=True).reshape(1, m)
cov_l = np.percentile(covs, q / 2 * 100, axis=0,
keepdims=True).reshape(m, m)
cov_u = np.percentile(covs, 100 - q / 2 * 100, axis=0,
keepdims=True).reshape(m, m)
cov_mu = mus.reshape(n_sim, m) - X.mean().to_numpy().reshape(1, m)
cov_mu = np.cov(cov_mu.T)
mu_l = pd.DataFrame(mu_l, index=[0], columns=cols)
mu_u = pd.DataFrame(mu_u, index=[0], columns=cols)
cov_l = pd.DataFrame(cov_l, index=cols, columns=cols)
cov_u = pd.DataFrame(cov_u, index=cols, columns=cols)
cov_mu = np.diag(np.diag(cov_mu))
cov_mu = pd.DataFrame(cov_mu, index=cols, columns=cols)
if au.is_pos_def(cov_l) == False:
cov_l = au.cov_fix(cov_l, method="clipped", threshold=1e-3)
if au.is_pos_def(cov_u) == False:
cov_u = au.cov_fix(cov_u, method="clipped", threshold=1e-3)
return mu_l, mu_u, cov_l, cov_u, cov_mu
def rp_optimization(self,
model="Classic",
rm="MV",
rf=0,
b=None,
hist=True):
r"""
This method that calculates the risk parity portfolio according to the
optimization model selected by the user. The general problem that
solves is:
.. math::
\begin{align}
&\underset{w}{\min} & & R(w)\\
&\text{s.t.} & & b \log(w) \geq c\\
& & & w \geq 0 \\
\end{align}
Where:
:math:`R(w)` is the risk measure.
:math:`b` is a vector of risk constraints.
Parameters
----------
model : str can be 'Classic' or 'FM'
The model used for optimize the portfolio.
The default is 'Classic'. Posible values are:
- 'Classic': use estimates of expected return vector and covariance matrix that depends on historical data.
- 'FM': use estimates of expected return vector and covariance matrix based on a Risk Factor model specified by the user.
rm : str, optional
The risk measure used to optimze the portfolio.
The default is 'MV'. Posible values are:
- 'MV': Standard Deviation.
- 'MAD': Mean Absolute Deviation.
- 'MSV': Semi Standard Deviation.
- 'FLPM': First Lower Partial Moment (Omega Ratio).
- 'SLPM': Second Lower Partial Moment (Sortino Ratio).
- 'CVaR': Conditional Value at Risk.
- 'CDaR': Conditional Drawdown at Risk of uncompounded returns.
rf : float, optional
Risk free rate, must be in the same period of assets returns.
Used for 'FLPM' and 'SLPM'.
The default is 0.
b : float, optional
The vector of risk constraints per asset.
The default is 1/n (number of assets).
hist : bool, optional
Indicate if uses historical or factor estimation of returns to
calculate risk measures that depends on scenarios (All except
'MV' risk measure). The default is True.
Returns
-------
w : DataFrame
The weights of optimum portfolio.
"""
# General model Variables
mu = None
sigma = None
returns = None
if model == "Classic":
mu = np.array(self.mu, ndmin=2)
sigma = np.array(self.cov, ndmin=2)
returns = np.array(self.returns, ndmin=2)
nav = np.array(self.nav, ndmin=2)
elif model == "FM":
mu = np.array(self.mu_fm, ndmin=2)
if hist == False:
sigma = np.array(self.cov_fm, ndmin=2)
returns = np.array(self.returns_fm, ndmin=2)
nav = np.array(self.nav_fm, ndmin=2)
elif hist == True:
sigma = np.array(self.cov, ndmin=2)
returns = np.array(self.returns, ndmin=2)
nav = np.array(self.nav, ndmin=2)
# General Model Variables
if b is None:
b = np.ones((1, mu.shape[1]))
b = b / mu.shape[1]
returns = np.array(returns, ndmin=2)
w = cv.Variable((mu.shape[1], 1))
rf0 = rf
n = returns.shape[0]
# MV Model Variables
risk1 = cv.quad_form(w, sigma)
returns_1 = af.cov_returns(sigma) * 1000
n1 = returns_1.shape[0]
risk1_1 = cv.norm(returns_1 @ w, "fro") / cv.sqrt(n1 - 1)
# MAD Model Variables
Y = cv.Variable((returns.shape[0], 1))
u = np.ones((returns.shape[0], 1)) * mu
a = returns - u
risk2 = cv.sum(Y) / n
# madconstraints=[a*w >= -Y, a*w <= Y, Y >= 0]
madconstraints = [a @ w <= Y, Y >= 0]
# Semi Variance Model Variables
risk3 = cv.norm(Y, "fro") / cv.sqrt(n - 1)
# CVaR Model Variables
alpha1 = self.alpha
VaR = cv.Variable((1, 1))
alpha = alpha1
X = returns @ w
Z = cv.Variable((returns.shape[0], 1))
risk4 = VaR + 1 / (alpha * n) * cv.sum(Z)
cvarconstraints = [Z >= 0, Z >= -X - VaR]
# Lower Partial Moment Variables
lpm = cv.Variable((returns.shape[0], 1))
lpmconstraints = [lpm >= 0, lpm >= rf0 - X]
# First Lower Partial Moment (Omega) Model Variables
risk6 = cv.sum(lpm) / n
# Second Lower Partial Moment (Sortino) Model Variables
risk7 = cv.norm(lpm, "fro") / cv.sqrt(n - 1)
# Drawdown Model Variables
X1 = 1 + nav @ w
U = cv.Variable((nav.shape[0] + 1, 1))
ddconstraints = [
U[1:] * 1000 >= X1 * 1000,
U[1:] * 1000 >= U[:-1] * 1000,
U[1:] * 1000 >= 1 * 1000,
U[0] * 1000 == 1 * 1000,
]
# Conditional Drawdown Model Variables
CDaR = cv.Variable((1, 1))
Zd = cv.Variable((nav.shape[0], 1))
risk10 = CDaR + 1 / (alpha * n) * cv.sum(Zd)
cdarconstraints = [
Zd * 1000 >= U[1:] * 1000 - X1 * 1000 - CDaR * 1000,
Zd * 1000 >= 0,
]
# Defining risk function
constraints = []
if rm == "MV":
if model != "Classic":
risk = risk1_1
elif model == "Classic":
risk = risk1
elif rm == "MAD":
risk = risk2
constraints += madconstraints
elif rm == "MSV":
risk = risk3
constraints += madconstraints
elif rm == "CVaR":
risk = risk4
constraints += cvarconstraints
elif rm == "FLPM":
risk = risk6
constraints += lpmconstraints
elif rm == "SLPM":
risk = risk7
constraints += lpmconstraints
elif rm == "CDaR":
risk = risk10
constraints += ddconstraints
constraints += cdarconstraints
# Frontier Variables
portafolio = {}
for i in self.assetslist:
portafolio.update({i: []})
# Optimization Process
# Defining solvers
solvers = [cv.ECOS, cv.SCS, cv.OSQP, cv.CVXOPT]
sol_params = {
cv.ECOS: {
"max_iters": 2000,
"abstol": 1e-10
},
cv.SCS: {
"max_iters": 2500,
"eps": 1e-10
},
cv.OSQP: {
"max_iter": 10000,
"eps_abs": 1e-10
},
cv.CVXOPT: {
"max_iters": 2000,
"abstol": 1e-10
},
}
# Defining objective function
objective = cv.Minimize(risk * 1000)
constraints += [b @ cv.log(w) * 1000 >= 1 * 1000, w * 1000 >= 0]
try:
prob = cv.Problem(objective, constraints)
for solver in solvers:
try:
prob.solve(solver=solver, **sol_params[solver])
except:
pass
if w.value is not None:
break
weights = np.array(w.value, ndmin=2).T
weights = np.abs(weights) / np.sum(np.abs(weights))
for j in self.assetslist:
portafolio[j].append(weights[0, self.assetslist.index(j)])
except:
pass
rp_optimum = pd.DataFrame(portafolio,
index=["weights"],
dtype=np.float64).T
return rp_optimum
