Hedge Fund Decomposition with Traditional and Alternative Investment Strategies

Peter Tsyurmasto
Posted on Jul 1, 2019

Introduction

During the 2008 financial crisis, institutional investors such as pension funds, insurance companies and banks suffered large losses. Hedge funds in particular had failed to provide financial strategies and returns uncorrelated to markets at a time when they were most needed. Over 10 years since then,  the natural question remains whether the financial industry is well-equipped for another crisis. The inherent complexity of financial products makes it challenging to effectively monitor the underlying risks of a portfolio, yet it is an essential task. In that spirit, we aim to create a simple tool that decomposes portfolio returns into investment strategies and explains the returns of hedge funds. The prototype of this tool can be found on ShinyApps.io, and the source code on GitHub.

Portfolio Decomposition

Portfolio decomposition to identify underlying risk factors is a standard approach for monitoring portfolio risk. This approach expresses our portfolio returns as a linear combination of factor returns with coefficients, a constant term and a residual vector. Coefficients (denoted beta) are exposures of portfolio to risk factors. The constant term (denoted alpha) shows whether the portfolio outperforms a basket of risk factors. Alpha > 0 indicates an outperforming portfolio while alpha < 0 indicates an underperforming portfolio.

Optimal coefficients can be found by minimizing the standard deviation between portfolio return and a risk factor basket – the concept known in finance as a tracking error. In this way, we decompose the variance of the portfolio into variances of individual factors and a remaining idiosyncratic (unexplained) variance component.

Data

We apply traditional and alternative investment strategies across multiple asset classes: equities, fixed income, FX, commodities, and volatility. Traditional strategies are defined as long-only investments (as involving the S&P500, US 10-year sovereign bonds, and so on), while alternative strategies are defined as a combination of long and short positions. The alternative strategies are further sub-divided into carry, value and momentum investment styles. Carry is earned when holding a high-yielding asset and financing it with a low-yielding asset. Value is associated with buying undervalued assets and selling overvalued assets based off a predefined fair value metric. Momentum involves buying assets that appreciate, and selling assets that depreciate, over a specified time horizon (e.g. 12-month). 

Hedge funds currently account for nearly $3 trillion in assets globally. They can be classified into four different styles: equity long/short, event-driven, global macro, and relative value. In this analysis, we use Credit Suisse hedge fund indices that aggregate performances of underlying monthly hedge fund returns within each style. We also study the daily returns of the Credit Suisse Liquid Alternative Beta indices.

Visualization with ShinyApp

As with recipes and their ingredients, hedge funds can be broken down into the building blocks of their underlying investment strategies. The portfolio decomposition tab in our app partitions the traditional and alternative investment strategies of each portfolio and allocates percentages to them. The chart below shows one example of factor exposures generated by the app for a Credit Suisse Lab daily hedge fund index.

The variance decomposition chart shows the percentages of portfolio risk attributed to each strategy and a remaining (idiosyncratic) component that cannot be explained by our factors. 

 

About Author

Peter Tsyurmasto

Peter Tsyurmasto

Dr. Peter Tsyurmasto has over 5 years working in financial industry applying statistical and machine learning models to solve complex financial problems. Peter earned a Ph.D. in operations research from University of Florida and Master's in Machine Learning...
View all posts by Peter Tsyurmasto >

Leave a Comment

No comments found.

View Posts by Categories


Our Recent Popular Posts


View Posts by Tags

2019 airbnb alumni Alumni Interview Alumni Spotlight alumni story Alumnus API Application artist aws beautiful soup Best Bootcamp Best Data Science 2019 Best Data Science Bootcamp Big Data Book Launch Book-Signing bootcamp Bootcamp Prep Bundles California Cancer Research capstone Career Career Day citibike clustering Coding Course Demo Course Report D3.js data Data Analyst data science Data Science Academy Data Science Bootcamp Data Scientist Data Scientist Jobs data visualization Deep Learning Demo Day Discount dplyr employer networking feature engineering Finance Financial Data Science Flask gbm Get Hired ggplot2 googleVis Hadoop higgs boson Hiring hiring partner events Industry Experts Job Jon Krohn JP Morgan Chase Kaggle lasso regression Lead Data Scienctist Lead Data Scientist leaflet linear regression Logistic Regression machine learning Maps matplotlib Medical Research meetup Networking neural network Neural networks New Courses nlp NYC NYC Data Science nyc data science academy NYC Open Data NYCDSA NYCDSA Alumni Open Data painter pandas Part-time Portfolio Development prediction Prework Programming PwC python python machine learning python scrapy python web scraping python webscraping Python Workshop R R language R Programming R Shiny r studio R Visualization R Workshop R-bloggers random forest recommendation recommendation system regression Scrapy scrapy visualization seaborn Selenium sentiment analysis Shiny Shiny Dashboard Spark Special Special Summer Sports statistics streaming Student Interview Student Showcase SVM Tableau TensorFlow Testimonial tf-idf Top Data Science Bootcamp twitter visualization web scraping Weekend Course What to expect word cloud word2vec XGBoost yelp