Navigating the Muddy Waters of Real Assets Benchmarking

6.18.2019

By TC Rolfstad, Vice President, Real Assets

In the world of private real assets, benchmarking represents a particularly challenging and complex task, as compared to traditional assets. Within certain areas of private real assets benchmarking is better defined, with industry-accepted indices and methodologies. Elsewhere, the industry is attempting to play catch-up as new investment options become available and plan sponsors evolve their portfolio structures and objectives. Parsing through all the nuances inherent in private real assets benchmarking would require exceeding the word count that we’ve allotted ourselves for this article – or, let’s be honest, that most people would be interested in reading. Still, these are crucial concepts to investing successfully in private real assets, so we have endeavored to relay a number of core principles that we believe are most relevant for institutional investors today.

When benchmarking private real asset portfolios, there are three primary elements that can influence returns: asset allocation, investment selection and market timing. Additionally, returns can be evaluated at three levels: the overall real assets portfolio level, by each real asset sub-sector (e.g., real estate, infrastructure, natural resources) and, finally, at the individual fund level (e.g., Manager A, Manager B). At each level, we place varying degrees of emphasis on the evaluation of the elements of return; therefore, the index by which we measure may differ by level.

But let’s not get too far ahead of ourselves. After all, before we unpack each of these levels, it’s important first to define the objectives of most real asset investors. Typically, investors commit capital to real assets to achieve some combination of the following: (i) portfolio diversification due to low levels of correlation with traditional asset classes, (ii) inflation hedging, (iii) potential compelling total returns, and (iv) yield. Unlike some alternative asset classes, most investors do not invest in real assets simply to achieve outperformance over the public markets.

Further, developing a diversified real assets portfolio requires careful attention to portfolio construction. You cannot simply invest in a single fund and be diversified across sub-sector and region, much less across risk type, vintage year and/or deal size. Building this framework requires identifying the appropriate market betas that, in aggregate, align with the overall portfolio objectives, followed by selecting the best-fit index. The final step involves identifying the individual fund investments that are capable of adding alpha within their specific betas.

There are four main types of benchmarks, each suited to a particular portfolio configuration:

  1. Private Market Indices (e.g., NCREIF Property Index)
    1. Best for evaluating diversified portfolios
  2. Private Market Peer Group Indices (e.g., Cobalt)
    1. Best for evaluating individual fund performance
  3. Economic Indicators (e.g., Consumer Price Index)
    1. Best for evaluating diversified portfolios
  4. Public Market Indices (e.g., DJ Brookfield Global Infrastructure Index)
    1. Best for evaluating diversified portfolios only if public market outperformance is an objective

 

At the overall real assets portfolio level, the primary goal of benchmarking is to evaluate your asset allocation decisions and underlying fund selections to determine whether they are achieving the top-line goals of the portfolio. At this level, market timing is only of minor consideration for most large institutional investors. And, just as no single fund is capable of gaining full diversification across real assets, neither is there one single index that fully captures the asset class.

Ultimately, the best index to utilize at the portfolio level is the one that aligns most closely with the overarching goal of the portfolio. If the goal is simply to outpace inflation, perhaps the Consumer Price Index is the right choice. If the goal instead is to try to generate total return and diversify away from traditional assets, then a private market index is likely most appropriate, or a hybrid index combining several private market indices. If the goal is to generate yield, the better comparison may be a fixed income index. Unless the goal is to outperform the public markets, a public market index is rarely the best or most relevant option for benchmarking private real assets.

Another factor to keep in mind when selecting a benchmark index ties back to the asset allocation decisions that originally prompted a commitment to real assets. Most institutional investors utilize a form of Modern Portfolio Theory that seeks to develop optimal portfolios using a model-based approach. These models typically rely on a set of assumptions for each asset class, and often those assumptions are based upon a reference index for that specific asset class. For purposes of continuity, it’s important to keep the reference index in mind as you select an overall real assets index, as the defining characteristics of the index that were on exhibit during the portfolio construction process are generally those desired to be present in the ensuing portfolio.

At the real asset sub-sector level, the goals are much the same as those of the overall real asset portfolio level where the focus is on asset allocation and investment selection. In this case, although index selection is often much more targeted, it is more intuitive, since sub-sector-specific index options exist. Most investors will select a private market index for evaluating performance at this level, though broad private peer group indices can be a useful option particularly in areas where private market indices are not available.

Finally, it comes down to evaluating performance at the individual fund level. At this level, we’re looking at performance relative to beta, but primarily we want to understand whether these individual investments have generated alpha. Given that most investments are far more nuanced at the individual fund level, it seems most practical to evaluate performance relative to peers using an appropriate private market peer group index. Peer groups are typically defined by a combination of factors including vintage year, geography and sub-sector; sometimes additional factors are taken into consideration. In this scenario, our goal is to understand whether a specific fund has generated top-quartile performance.

We often are asked whether a set return premium (e.g., +300 bps) should be added to private market indexes for benchmarking private real assets, particularly when the strategy being pursued by a client has a higher return and risk profile than the index. This approach is common across traditional private equity, as the premium over a public market index most often is intended to represent the return premium required for the lack of liquidity in the space and can reflect the long-term outperformance of the asset class relative to public markets. However, and as we mentioned, it’s more the exception than the rule that investors are committing to private real assets simply to outperform the public markets. While this potential outcome is appreciated, it’s often farther down the list of reasons for investing in private real assets. Instead, the logic often offered for using a premium over private real asset indices is to account for leverage or strategy differences between the index and the expected portfolio. While this attempt sounds attractive in principle, we believe it is difficult to objectively determine what static level of return premium is best for capturing a dynamic portfolio - is it +300 bps, +500 bps or more? Because of the lack of objective support, we often find investors circling around years later to analyze relative performance; if underperformance occurs, those investors may very well have difficulty determining whether that is due to true underperformance or poor index composition. Thus, we typically avoid the use of set return premiums because of their seemingly arbitrary application and questionable subsequent analysis takeaways. This is one area where the available options of benchmarking private real assets simply fall short.

Investors looking to benchmark a private real assets portfolio need to first identify their portfolio goals and objectives, and then determine what elements of performance are most relevant for measurement at the overall portfolio, sub-sector and individual fund levels. Finally, investors should focus on selecting specific benchmarks that are best geared to analyze those performance elements.


5/21/2019
Choose Your Own Downturn
« previous
6/20/2019
Co-Investing: The Struggle is Real
next »