ashkonf / quantitativehedging

Quantitative Hedging

Contents

• Overview
• Dependencies
• Usage
• Example
• License
• Links

Overview

The Quantitative Hedging repository provides an easy way to hedge a stock using a basket of other stocks which collectively behave as a hedge against the desired stock. The repo is intended for two types of users: (1) market makers who need to offset the risk derived from undesired inventory and (2) quantitative researchers who need to identify factors or replicate studies involving the performance of a security, portfolio, or hedge fund.

Dependencies

Trading Baskets requires the following libraries:

• pandas
• cvxopt
• numpy

Install these libraries using pip with requirements.txt:

pip install -r requirements.txt

Usage

This repo exports one public functions (in hedge.py) build_basket which builds the basket of stocks intended to hedge a desired stock.

build_basket()

Use build_basket() to compose a hedging basket from (1) the stock to be hedged (hedged_ticker_symbol) and (2) the ticker symbols to consider including in the hedging basket (basket_ticker_symbols):

 basket(build_basket, basket_ticker_symbols)

The basket() arguments are as follows:

Name	Type	Description	Optional?	Sample Value
hedged_ticker_symbol	str	The stock-ticker name of the stock to be hedged.	No	"AAPL"
basket_ticker_symbols	list	A list of ticker symbols the library will consider including in the hedging basket.	No	["GOOG", "MSFT", "NFLX", "AMZN", "FB"]

Example

The code below shows how to hedge a stock. The code defines APPL (Apple) as the stock to hedge, a list of stocks to consider using as part of the hedge (GOOG, MSFT, NFLX, AMZN, and FB), and composes a corresponding hedge basket for AAPL.

from hedge import build_basket

hedged_ticker_symbol = "AAPL"
basket_ticker_symbols = ["GOOG", "MSFT", "NFLX", "AMZN", "FB"]
print("Hedge for %s:" % hedged_ticker_symbol)
print(build_basket(hedged_ticker_symbol, basket_ticker_symbols))

This will produce the following hedging basket:

{'AAPL': 0.2614353523521262, 'FB': 0.1921680128468791, 'AMZN': 0.5463966348009947}

i.e. AAPL with weight 27%, FB with weight 19%, and AMZN with weight 54%.

A snippet like this can be incorporated in any Python application.

License

Trading Baskets is licensed under the Apache License 2.0.

Links

• Apache License 2.0
• CVXOPT
• NumPy
• pandas