pyfinance

pyfinance is a Python package for investment management and analysis of security returns. It complements existing quantitative-finance packages such as pandas-datareader with a focused, batteries-included statistical toolkit.

Status

• Latest release: 2.0.2
• Python: 3.10, 3.11, 3.12, 3.13, 3.14
• License: MIT

Installation

uv add pyfinance

Or with pip:

pip install pyfinance

Modules

Module	Description
pyfinance.returns	Statistical analysis of financial time series via the CAPM framework. TSeries / TFrame are Pandas subclasses that add performance statistics: annualized return/vol, Sharpe, Sortino, drawdown, capture ratios, alpha/beta, Information Ratio, and so on.
pyfinance.ols	Ordinary least-squares regression. OLS, RollingOLS (NumPy-backed), and PandasRollingOLS (Pandas-indexed wrapper).
pyfinance.options	Vectorized Black-Scholes-Merton valuation, Greeks, and implied volatility via BSM. Option strategies (Straddle, BullSpread, Butterfly, Condor, variants).
pyfinance.general	General-purpose computations: active share, amortization schedules, best-fit distribution, PCA on returns, portfolio simulation, tracking-error optimization, VIF.
pyfinance.datasets	A small set of dataset loaders: load_13f (SEC EDGAR), load_industries (Ken French), load_rates (FRED H.15), load_rf (3-month T-bill total-return series).
pyfinance.utils	Frequency conversion, rolling-window construction, one-hot encoding, random ticker/weights generation, availability reporting.

Quick tutorial

TSeries — performance statistics

import numpy as np
import pandas as pd
from pyfinance import TSeries

rng = np.random.default_rng(444)
s = rng.standard_normal(400) / 100 + 0.0008
idx = pd.date_range(start="2016-01-01", periods=len(s))
ts = TSeries(s, index=idx)

ts.anlzd_ret()       # annualized geometric return
ts.anlzd_stdev("D")  # annualized stdev of returns
ts.max_drawdown()    # worst peak-to-trough decline
ts.sharpe_ratio()    # Sharpe, annualized

BSM — Black-Scholes-Merton options pricing

from pyfinance.options import BSM

op = BSM(S0=100, K=100, T=1, r=0.04, sigma=0.20)
op.value()          # European call value
op.delta()          # Greeks
op.implied_vol(10)  # implied vol at a target price

# Vectorized across arrays of strikes.
import numpy as np
ops = BSM(S0=100, K=np.arange(100, 110), T=1, r=0.04, sigma=0.20)
ops.value()

OLS — regression

from pyfinance.ols import OLS

# y: 1-d array; x: 2-d array of explanatory variables.
model = OLS(y=y, x=x)
model.beta, model.alpha
model.rsq_adj, model.fstat, model.std_err

datasets.load_13f — SEC EDGAR 13F parser

SEC EDGAR requires a descriptive User-Agent (policy). Supply one:

from pyfinance import datasets

url = (
    "https://www.sec.gov/Archives/edgar/data/1040273/"
    "000108514617001787/form13fInfoTable.xml"
)
df = datasets.load_13f(url, user_agent="Jane Doe (jane@example.com)")

Development

This repo is managed with uv:

git clone https://github.com/bsolomon1124/pyfinance
cd pyfinance
uv sync               # install pinned runtime + dev deps
uv run pytest         # run tests
uv run ruff check     # lint
uv run ruff format    # format
uv build              # build sdist + wheel

Type checking uses ty:

uv run ty check

CI runs these across Python 3.10–3.14 on every push; see .github/workflows/ci.yml.

Changelog

See CHANGELOG.md. Notably, pyfinance 2.0 dropped load_factors, load_shiller, and load_retaildata because their upstream sources had drifted past reasonable repair. See the 2.0 release notes for details and migration guidance.

License

MIT — see LICENSE.