Hobbled by Benchmarks

  • Mike Aked, Rob Arnott, Omid Shakernia, Jonathan Treussard
  • Journal of Portfolio Management
  • A version of this paper can be found here
  • Want to read our summaries of academic finance papers? Check out our Academic Research Insight category.

What are the research questions?

Despite Peter Bernstein’s suggestion in 2003 that adherence to a fixed and undiversified policy portfolio is dangerous, benchmarks ( the most used of which is the 60/40) are as popular today as they were 15 years ago. The authors study the following research question:
  1. Is a broad investment opportunity set ( an un-levered and long-only version of a “value, momentum and carry” TAA system applied to a diverse universe of asset classes) a powerful source of incremental investment return?

What are the Academic Insights?

By studying three multi-asset class universes ( 2 assets US-based; 4 assets Globally-based; 15 assets Global diversified set) over the period from 1980 to 2016 and using simple model set-ups ( i.e. equal weighting of active TAA bets), the authors find the following:

  1. YES- An investor with a 2% peer-group active risk target could have added economically significant (annualized risk-adjusted alpha of 122 bps) returns in an un-levered and long only “diversified” portfolio by adding value, momentum and carry simple signals ( simplicity is used to avoid data mining). Additionally, the authors show that diversifying by strategy ( the combination of the different signals) or by the universe ( moving to the 15 asset class portfolio) coincides with a strong upward trend in alphas and t-stats.

Why does it matter?

While this study is simplistic, the authors hope to drive home the following message: “Please embrace diversification in your quest for long-term investment success.”

The Most Important Chart from the Paper:


Abstract

In 2003, at the NMS Endowments and Foundations Conference, Peter Bernstein suggested that policy portfolios are overused, leading to excessive tracking-error constraints. In a 2004 article in this journal, Rob Arnott showed that the 20 largest U.S. corporate pension funds were willing to accept 12% annual volatility in total return, 15% volatility relative to liabilities, but only 2.5% tracking error relative to peers. Today, the use of policy portfolios is as dominant as it was 15 years ago; policy portfolios and their respective benchmarks continue to be a largely home-centric 60/40 allocation. In the view of the authors, there should be no debate on the benefits of broadly (i.e., globally) diversified policy portfolios. Broad diversification is a requirement for actively adding value over time. The authors believe that long-only investors can translate the lightly correlated sources of excess returns from the long–short space to their portfolios. They hope that investors will hear their plea: Please diversify your benchmarks.