blotter: a financial blotter for Python

Description

blotter is a python object for representing a financial blotter. The main goal of blotter is to provide a decoupled way to manage historical holdings and track performance when backtesting a trading strategy which trades FX and Futures. blotter is not intended for calculating alpha's, simulating execution or performing portfolio optimization but rather serves as an endpoint to book the price of executed trades and provide marked to market holdings and PnL.

Main Features

• Handle open and closed PnL calculations across multiple currencies
• Ability to specify base currency for marked to market holdings calculations, automated PnL sweeps and margin charges
• Accounts for commissions
• Allows for trading instruments denominated in different currencies
• Automatically manage daily interest payments, margin charges, PnL calculations and PnL sweeps or manage manually
• Get intraday mark to market holdings denominated in base currency

Install

You can pip install this package from github, i.e.

pip install git+git://github.com/matthewgilbert/blotter.git@master

the package is not yet available on Pypi.

Requires

• pandas >= 0.18
• numpy

For building the documentation sphinx is also required and performance tests use asv

Documentation

Online documentation can be found at https://matthewgilbert.github.io/blotter/ The documentation is built using Sphinx and its source is available in doc/

Tests

Unit tests are located in blotter/tests/ and can be run using make test