Stock Exchange

Benchmarks and pension funds, we can do better!

Until the late seventies equity investment portfolios were primarily judged on their overall efficiency. Benchmarks emerged later. The purpose of the benchmark is to measure the performance of the fund manager relative to a representative index. Indices whose weights are proportional to the market value of a company were chosen on the basis of scientific arguments. In this short article we argue that investors such as pension funds will benefit from taking a closer look to and call into question the composition of their benchmarks. We state that pension funds will benefit from the use of ‘smart’ benchmarks.

Market Value-weighted Indices

Market value-weighted indices are compiled by index providers. They decide which listed companies are included in the index. Subsequently, the index weight is determined by the value of the company on the stock exchange. This allows investors to indirectly affect the weight of a company in the index. The index weight of stocks that rise faster than the market will increase and the weight of stocks that lag behind the market will decrease. For example, the price of Apple shares rose from $95 in 2008 to $700 in 2012. Simultaneously the index weight in the world index grew from 0.45 percent to 2.45 percent! The advantage of index weights that automatically move along with the price movements is that weights automatically adjust. There are no transactions necessary on the stock exchange, as long as the index does not change in composition. However, more recently the disadvantages of market capitalization weighted indices were identified:

* Sub-optimal diversification

In general, marked capitalization-weighted benchmarks do not necessarily have optimal diversification. Stocks are simply selected by the index provider because of their market value. Expected returns, risks and correlations do not play part in the selection at all. The potential benefits of a smart way to deal with correlations and risks are not used.

* Risk concentration

The name of the index remains the same, but the risk of the benchmark changes every day. The weights of stocks, regions and industries fluctuate over time due to the focus on market values. In other words, an investor who follows a market value-weighted index takes risks that he does not actively manage himself. He leaves choices about certain risks to other investors in the market.

New Insights

Both scientists and asset managers recently made substantial progress in their thoughts about  market value-weighted benchmarks. New insights have led to the creation of investment strategies that react less on the market capitalization of stock. The main developments are the emergence of theme benchmarks based on specific stock characteristics other than the market value, just like a Minimum Risk benchmark. Recent academic research also provides evidence that there are attractive alternatives to market value-weighted benchmarks.

Theme Benchmark Portfolios Based on Stock Characteristics

The first logical step in the release of “market value-weighted index”-thinking is a  benchmark weight based on other enterprise or stock characteristics than the market value. The book value of equity, revenue and equity beta are enterprise characteristics that can be eligible. The great advantage of this weighting scheme is that emotions on stock markets resulting in daily price movements influence the composition of these benchmark portfolios less. The investment strategies behind these benchmarks are systematic and do not require return forecasts. Benchmark portfolios in which the weights are based on the book value of a company are also known as ‘Fundamental Indexation’. Just like market capitalization weighted benchmarks, using book values for benchmarking, will make people invest more in larger companies than in smaller companies. The realized returns on ‘Fundamental Indexation’ benchmarks and market capitalization weighted benchmarks can vary considerably, there is a high tracking error. ‘Fundamental Indexation’-strategies had better returns historically. A disadvantage however is that a theoretical underpinning for such a weighting scheme is missing in our opinion.

Benchmark Portfolios Based on Minimum Risk

Minimum Risk benchmark portfolios let go the use of market value-weighted benchmarks. A Minimum Risk  portfolio is a portfolio that –on the basis of one or more risk models- has the lowest expected risk of all potential equity portfolios. Such portfolios have been known since the pioneering study of Markowitz in 1959. During the last few years these portfolios have become increasingly popular. The advantage is that yield predictions are not required. The portfolio can be assembled without any expectations of the expected returns. This is a great feature when there is much uncertainty about the expected return of individual stocks, while there is more certainty about the risk of a well-diversified equity portfolio.

Minimum Risk portfolios are systematic strategies and are constructed based on optimization routines, risk models and transaction cost models.  Besides striving for the lowest possible risk, further demands on Minimum Risk portfolios are required in practice too. The maximum weight of each share is usually limited, as well as sector and region weights. Also, the sales and style influences are limited. These additional requirements are intended to create a neutral yield, alternative to market capitalization-weighted investing. In practise, Minimum Risk portfolios have a twenty to thirty percent lower volatility than market value-weighted indices.

Empirical analyzes based on historical data shows that the Minimum Risk strategy yields higher returns than market value-weighted benchmarks. Due to lower risk and higher average returns Minimum Risk strategies show attractive Sharpe ratios for most equity universes: the world, the United States, Europe and emerging markets. These historical analyzes show that Minimum Risk portfolios are more efficient than traditional market value-weighted benchmarks. Obviously, the risk reduction applies to the time the portfolio is made (ex ante), but also appears to apply to the actual risks (ex post). The volatility of the portfolio is generally reduced by twenty to fifty percent. The biggest reduction in risk will be achieved in periods of high market volatility such as the credit crunch. This is exactly the time when pension funds welcome risk reduction the most.

Smart Benchmarks and Pension Funds

Perhaps the most important development is that smart benchmarks become accepted by pension fund managers in The Netherlands. In fact The Netherlands is, together with the Scandinavian pension funds, at the leading edge of implementing the new benchmark-thinking. The idea is that pension investments can be more efficient, but without the use of complex instruments and at lower cost.

Several Dutch pension funds now invest a –significant- portion of their equity portfolio through this clever benchmarks. Dutch financial institutions actively participate in questioning the benchmark paradigm through empirical research. The historically strong orientation of Dutch universities on the field of econometrics and statistics contribute to this.

Gerben de Zwart, PhD and Ronald van Dijk, PhD are pension investor at APG as well as Head of Quantitative Equity Research and Managing Director Developed Market Equities.

[1] This article is an update and summary of the article ‘Actief omgaan met Risico’ published in the VBA Journal.

This article is translated from Dutch to English by the Editorial Board of Asset | Accounting & Finance.