squarequant 0.2.3

A Python package for financial risk metrics and stock data analysis

SquareQuant

SquareQuant is a comprehensive Python package for financial risk metrics and stock data analysis. It provides a suite of tools for quantitative finance, including risk metrics calculation and data retrieval from multiple sources.

Features

• Financial Data Access: Multiple data sources including YFinance and Theta Data
• Extensive Risk Metrics: Comprehensive set of traditional and advanced risk measures
• Modular Design: Organized into specialized modules (data, var, etc.)
• Flexible API: Consistent function-based and object-oriented interfaces

Installation

pip install squarequant

Quick Start

import squarequant as sq
import pandas as pd
import matplotlib.pyplot as plt

# Configure data download
config = sq.DownloadConfig(
    start_date="2020-01-01",
    end_date="2023-12-31",
    interval='1d',
    columns=['Close'],
    source="yfinance"  # Or "theta" for Theta Data
)

# Download data
data = sq.download_tickers(['AAPL', 'MSFT', 'GOOGL'], config)
data = pd.DataFrame({
    'AAPL': data['AAPL_Close'],
    'MSFT': data['MSFT_Close']
})

# Calculate volatility
vol = sq.vol(data=data, assets=['AAPL', 'MSFT', 'GOOGL'], use_returns=False, window=21)

# Calculate Sharpe ratio
sharpe = sq.sharpe(data=data, assets=['AAPL', 'MSFT', 'GOOGL'], use_returns=False, window=252)

# Visualize results (using matplotlib directly)
plt.figure(figsize=(12, 6))
plt.plot(vol)
plt.title('21-Day Rolling Volatility')
plt.legend(vol.columns)
plt.show()

Available Risk Metrics

SquareQuant provides a comprehensive set of risk metrics, including:

Metric	Function	Description
Sharpe Ratio	sq.sharpe()	Risk-adjusted return using standard deviation
Sortino Ratio	sq.sortino()	Risk-adjusted return using downside deviation
Volatility	sq.vol()	Standard deviation of returns
Maximum Drawdown	sq.mdd()	Largest peak-to-trough decline
Value at Risk	sq.var()	Maximum expected loss at given confidence level
Conditional Value at Risk	sq.cvar()	Expected loss beyond the VaR threshold
Semi-Deviation	sq.semidev()	Downside volatility below target return
Average Drawdown	sq.avgdd()	Mean of all drawdowns in a period
Ulcer Index	sq.ulcer()	Square root of mean squared drawdown percentage
Mean Absolute Deviation	sq.mad()	Average absolute deviation from mean
Entropic Risk Measure	sq.erm()	Risk measure based on information theory
Entropic Value at Risk	sq.evar()	VAR using entropy concepts
Conditional Drawdown at Risk	sq.cdar()	Expected drawdown beyond threshold

Data Sources

SquareQuant now supports multiple data providers:

YFinance

The default data provider for historical price data:

config = sq.DownloadConfig(
    start_date="2020-01-01",
    end_date="2023-12-31",
    source="yfinance"  # Default
)

Theta Data

A new alternative data provider with the same interface:

config = sq.DownloadConfig(
    start_date="2020-01-01",
    end_date="2023-12-31",
    source="theta"
)

Advanced Usage Examples

Analyzing Multiple Risk Metrics

import squarequant as sq
import matplotlib.pyplot as plt

# Download data
config = sq.DownloadConfig(
    start_date="2018-01-01",
    end_date="2023-12-31",
    interval='1d'
)
data = sq.download_tickers(['SPY', 'QQQ'], config)

# Calculate multiple risk metrics
vol = sq.vol(data=data, assets=['SPY', 'QQQ'], use_returns=False, window=21)
mad = sq.mad(data=data, assets=['SPY', 'QQQ'], use_returns=False, window=252)
mdd = sq.mdd(data=data, assets=['SPY', 'QQQ'], use_returns=False, window=252)
ulcer = sq.ulcer(data=data, assets=['SPY', 'QQQ'], use_returns=False, window=252)

# Visualize comparison (using matplotlib directly)
plt.figure(figsize=(12, 8))
plt.subplot(3, 1, 1)
plt.plot(vol['SPY_Vol'])
plt.title('SPY Volatility')

plt.subplot(3, 1, 2)
plt.plot(mdd['SPY_MDD'])
plt.title('SPY Max Drawdown')

plt.subplot(3, 1, 3)
plt.plot(ulcer['SPY_Ulcer'])
plt.title('SPY Ulcer Index')

plt.tight_layout()
plt.show()

Portfolio Risk Analysis

import squarequant as sq
import pandas as pd

# Example portfolio weights
weights = pd.Series({'AAPL': 0.3, 'MSFT': 0.3, 'AMZN': 0.2, 'GOOGL': 0.2})

# Download data
config = sq.DownloadConfig(start_date="2020-01-01", end_date="2023-12-31")
data = sq.download_tickers(weights.index.tolist(), config)

# Calculate portfolio risk measures
portfolio_var = sq.var(
    data=data,
    use_returns = False,
    assets=weights.index.tolist(),
    confidence=0.95,
    window=252,
    holding_period=10
)

# Print results
print(f"Latest Portfolio VaR (95%): {portfolio_var.iloc[-1].values[0]:.4f}")

Using Advanced Risk Measures

import squarequant as sq

# Download data
config = sq.DownloadConfig(start_date="2018-01-01", end_date="2023-12-31")
data = sq.download_tickers(['SPY'], config)

# Calculate Entropic Risk Measure with different risk aversion parameters
erm1 = sq.erm(data=data, use_returns=False, assets=['SPY'], z=0.5, confidence=0.95)
erm2 = sq.erm(data=data, use_returns=False, assets=['SPY'], z=1.0, confidence=0.95)
erm3 = sq.erm(data=data, use_returns=False, assets=['SPY'], z=2.0, confidence=0.95)

# Compare with traditional VaR and CVaR
var = sq.var(data=data, use_returns=False, assets=['SPY'], confidence=0.95)
cvar = sq.cvar(data=data, use_returns=False, assets=['SPY'], confidence=0.95)

print(f"Latest ERM (z=0.5): {erm1.iloc[-1, 0]:.4f}")
print(f"Latest ERM (z=1.0): {erm2.iloc[-1, 0]:.4f}")
print(f"Latest ERM (z=2.0): {erm3.iloc[-1, 0]:.4f}")
print(f"Latest VaR (95%): {var.iloc[-1, 0]:.4f}")
print(f"Latest CVaR (95%): {cvar.iloc[-1, 0]:.4f}")

Library Structure

The package has been reorganized into specialized modules:

• squarequant.data: Data retrieval functionality (YFinance, Theta Data)
• squarequant.core: Core risk metrics and calculations
• squarequant.var: Value at Risk and related risk measures
• squarequant.monte_carlo: Monte Carlo simulations (alpha version)

Documentation

For detailed documentation on each function and class, please see the official documentation.

License

MIT License

Credits

SquareQuant was developed to provide financial analysts, quantitative researchers, and portfolio managers with a comprehensive toolkit for risk assessment and performance analysis.