Skip to content

Expected Returns: FactorAnalytics

Jump to bottom Edit New page
Justin M. Shea edited this page Feb 19, 2021 · 8 revisions

Background

As witnessed by recent events, the lack of portfolio diversification and risk control can severely impact the financial goals and long term plans for individual investment accounts, retirement accounts, University Endowment funds, and Municipal Pension funds alike. And as recent examples of speculative booms and busts have revealed (GameStop), investors sometime exhibit a lack of diversification among their securities holdings, and investments are not as safe as they initially seem.

In this project, you shall explore and implement several investment strategies in R, as inspired by one of the most interesting investment references on the topic Expected Returns: An Investors Guide to Harvesting Market Rewards by Antti Ilmanen.

From the Description;

This comprehensive reference delivers a toolkit for harvesting market rewards from a wide range of investments. Written by a world-renowned industry expert, the reference discusses how to forecast returns under different parameters. Expected returns of major asset classes, investment strategies, and the effects of underlying risk factors such as growth, inflation, liquidity, and different risk perspectives, are also explained. Judging expected returns requires balancing historical returns with both theoretical considerations and current market conditions. Expected Returns provides extensive empirical evidence, surveys of risk-based and behavioral theories, and practical insights.

Related work

Your objective will be to reproduce key approaches suggested by the text and test performance on current market conditions with R. You will use functions found in popular R in finance packages such as FactorAnalytics, PerformanceAnalytics and PortfolioAnalytics. But you will also need to write functions of your own to streamline workflows and implement solutions. While these packages are excellent and widely used, there are gaps in the workflows involved in constructing portfolio management strategies we'd like to fill.

Details of the ExpectedReturns project & Impact

Mentors will guide your understanding of the topic, support your use of best practices in software development for quantitative finance using R, and provide market data for validating these approaches.

Ultimately, this work will be organized into an open source R package. It will complement the text and provide data, functions, and reproducible examples to guide academics, practitioners, and hobbyists in the R community in applying the work to their own research or portfolio management endeavors.

Students engaged in this project will obtain a deeper understanding of:
i) Data Science applications in finance
ii) Quantitative Analysis of active portfolio management

Areas of Interest

Approaches to Dynamic asset weighting

Value-oriented equity selection, chapter 12

Commodity Momentum and trend following, Chapter 14

Data

Most underlying data series are extracted from Bloomberg, including MSCI Barra’s equity indices, Barclays Capital and other banks’ bond indices, and S&P GSCI commodity futures indices. Other key sources include Kenneth French’s, Robert Shiller’s, and AQR's websites.

To create long data histories for major asset classes, the author concatenated best quality recent data and best available older data series. Most exhibits display total returns denominated in U.S. dollars; but some exhibits show real (inflation-adjusted) returns or excess returns over cash or over maturity/duration-matched Treasuries.

We envision the following steps for this project:

  • Skim the text above
  • Get familiar with the ExpectedReturns project.
  • Refactor, document, and unify existing functions, adding new ones as needed.
  • Refactor case studies in existing vignettes and unify them with the FactorAnalytics R package functions.
  • Convert existing vignettes to static vignettes
  • Add Unit tests using the tinytest R package
  • Submit PRs for finished vignettes for inclusion in the FactorAnalytics package

Mentors

Student-developer

Tests

Students, please do one or more of the following tests before contacting the mentors above. We encourage work on Linux Debian-based distributions.

  1. Easy: Begin by downloading and building the ExpectedReturns and FactorAnalytics packages locally. Work through, and list any build errors or issues you encounter on install.
library(remotes)
install_github("JustinMShea/ExpectedReturns")
install_github("braverock/FactorAnalytics")

  1. Intermediate: Locate the expected-returns-replications.Rmd file in the vignettes directory. Refactor sections of this vignettes to replace functions from the plm package with the fitFfm or fitFfmDT functions associated with the FactorAnalytics package. This may include debugging upstream issues with merging data series, as well as reformatting data to match requirements of the new function arguments.

  2. Harder: Reflect on the steps above. How do you interpret the results of the new functions? In addition, was there any repetitious code in the vignette that may be written as a function for future use? If so please include it as an example. What data transformations or models might have benefited from writing unit tests? Please include examples for these as well.

Solutions of tests

Students, please post a link to your test results here.

  • EXAMPLE STUDENT 1 NAME, LINK TO GITHUB PROFILE. (Email your test results to mentors)

References

Ilmanen, Anti. 2011. “Expected Returns.” John Wiley & Sons Ltd. ISBN: 978-1-119-99072-7

Please refrain from editing this Footer. If you want to edit this page, please scroll to the top of the web page and click the Edit button in the upper right.

Add a custom sidebar
Clone this wiki locally