Value Added By Large Institutional Investors Between 1992-2013

Alexander D. Beath, PhD

CEM Benchmarking Inc.

372 Bay Street, Suite 1000, Toronto, ON, M5H 2W9

Can large institutional investors beat the market and deliver added value above and beyond their benchmarks? We answer this question using a massive data set comprised of 6,666 samples drawn from a global set of defined benefit pension plans along with a handful of sovereign wealth funds and buffer funds spanning 1992?2013. Gross of investment management expenses, funds deliver 58 basis points of value added. Net of investment management expenses, funds deliver 16 basis points of value added. A deep regression analysis indicates that beating the market is rooted in active asset management paired with cost savings gained through scale and managing assets in?house.

Value Added By Large Institutional Investors Between 1992?2013 – Introduction

The primary way pension funds quantify their performance remains the tried and true method of constructing a total fund policy benchmark based on asset class policy weights and comparing their performance to it. While some in the industry see the method as being outdated, the benefits remain clear; fund performance is easily separated into two parts, a return that can be attributed to active management (i.e., alpha) and a return attributed to the market (i.e., beta). As the method is available to all, even those who feel it imperfect should perform the exercise to answer the simple question; did you get what you paid for?

A widely held academic belief is that the result of the exercise should show that active investors have, on average, no advantage over passive investors (i.e., alpha is zero). This view of markets is rooted in the efficient market hypothesis [1,2]. A problem with testing the hypothesis is that the separation between alpha and beta is not always clear; where one set of benchmarks demonstrates a non?zero alpha, another set can almost always be found that shows that the alpha is zero.

At CEM Benchmarking we are in the unique position of being able to provide answers to the question of whether pension funds are able to add value and beat the market. Not only can we definitively answer the question of whether it is possible, we can also quantify to a large degree how these institutional investors do it. What advantages do they have? Where have they added value? Is the value added really alpha, or is it beta in disguise? The answer in short is, yes, pension funds have added value. Gross of investment costs, the value added by pension funds over their policy return averaged 58 basis points. The key question though is whether it was worth it. Does the result justify the added costs and risk of active management? The answer here is again, yes, as net of investment costs the value added is 16 basis points1, on average. However, with three quarters of the performance chewed up by investment costs, it is clear why measuring and managing costs remains a critical element of pension fund management.

The CEM database

For over 25 years CEM Benchmarking has been collecting detailed information from pension plans and other large institutional investors on their asset allocation, benchmark returns, gross returns, investment costs, and by consequence net returns. The database is global, with participation from funds in Australia, Canada, Korea, the Netherlands, Sweden, the U.S., and the U.K. along with many others, and more than 1,000 unique funds have participated in it at one time or another. The database skews towards large funds, with an average size in 2013 of about $19.5 billion USD in assets under management (AUM). For the results presented here, funds are for the most part defined benefit (DB) pension plan, both public and corporate, along with a small number of sovereign wealth funds and buffer funds which share characteristics with their large DB counterparts. In Table 1 we summarize the characteristics of funds by year appearing in this study.

Value Added

A useful feature of the CEM database is the bias free nature of the data with respect to returns [3]. When funds benchmark with CEM they are benchmarking their investment costs. As part of the investment cost benchmarking process, investment performance metrics are provided as well in order to answer the question “do you get what you pay for?” which is increasingly top?of?mind for stakeholders. So while funds enter and exit the database over time, their motivation for doing so is unrelated to investment performance which explains why performance biases typically seen in other databases (e.g., hedge funds, mutual funds, etc., which suffer from survivorship bias [2]) is absent in the CEM database.

Total fund value added, net and gross

A standard part of the CEM survey asks funds to provide their calendar year investment return along with their total fund policy benchmark return. Total fund benchmark return is a time weighted average of asset class policy weights and asset class benchmark returns. Gross value added is the difference between total fund return, before netting of investment costs, and the policy return. Net value added equals gross value added less investment costs. Spanning 1992?2013, there are 6,666 data points, more than sufficient to draw firm conclusions regarding fund performance.

Value Added

See full PDF below.