Riskfolio-Lib

7.2.1 verified Sat May 09 auth: no python

Riskfolio-Lib is a Python library for portfolio optimization and quantitative strategic asset allocation. It provides tools for mean-variance, Black-Litterman, risk parity, and other advanced portfolio methods. Current version is 7.2.1, requiring Python >=3.9. The library is actively maintained with a regular release cadence.

pip install riskfolio-lib

Common errors

error ModuleNotFoundError: No module named 'riskfolio' ↓

cause Importing with hyphen instead of underscore: 'riskfolio-lib' is the package name, but the module is 'riskfolio'.

fix

Use 'import riskfolio' after 'pip install riskfolio-lib'.

error KeyError: 'prices' ↓

cause Portfolio class in v7+ does not accept 'prices' as a parameter.

fix

Pass 'returns' (DataFrame of returns) instead of 'prices' to Portfolio.

error cvxpy.error.SolverError: Solver not found (ECOS, SCS, etc.) ↓

cause cvxpy is installed but no solver backend is available.

fix

Install a solver: e.g., 'pip install cvxopt'.

error ValueError: The covariance matrix is not positive semidefinite ↓

cause Data contains NaN or insufficient observations causing non-psd covariance.

fix

Drop or interpolate NaN values; use method='ledoit' in covariance estimation: Port = Portfolio(returns, method_cov='ledoit').

Warnings

breaking In version 7.0+, the Portfolio class no longer accepts 'prices' directly; use 'returns' or 'mu' and 'cov' explicitly. ↓

fix Convert prices to returns with .pct_change().dropna() before passing to Portfolio.

deprecated The function 'riskfolio.MeanRisk()' is deprecated. Use 'Portfolio.optimization()' with risk measure parameters. ↓

fix Switch to Portfolio object and its optimization method.

gotcha The library requires cvxpy with a compatible solver (e.g., ECOS, SCS). Missing solver may cause silent fallback or errors. ↓

fix Install cvxpy and optionally 'cvxopt' or 'scipy' for alternative solvers.

Imports

Portfolio
wrong
```
import Portfolio from riskfolio-lib
```
correct
```
from riskfolio import Portfolio
```
Hyphen in library name, but Python module uses underscore; import as 'riskfolio'.
HRP (Hierarchical Risk Parity)
wrong
```
from riskfolio.lib import HRP
```
correct
```
from riskfolio import HCPortfolio
```
HRP is implemented in HCPortfolio class, not a separate module.

RiskParity

wrong

from riskfolio.optimization import RiskParity

correct

from riskfolio import RiskParity

RiskParity is in the top-level namespace.

Quickstart

Basic mean-variance optimization to maximize Sharpe ratio using historical returns.

import numpy as np
import pandas as pd
from riskfolio import Portfolio

# Sample data
ereturns = pd.DataFrame(np.random.randn(100, 4), columns=['Asset1','Asset2','Asset3','Asset4'])

# Portfolio object
port = Portfolio(returns=ereturns)

# Mean-variance optimization
w = port.optimization(model='Classic', rm='MV', obj='Sharpe', hist=True)
print(w)