Asset Allocation: Risk Models For Alternative Investments

Updated on

Asset Allocation: Risk Models For Alternative Investments by Niels Pedersen, Sébastien Page, CFA, and Fei He, CFA, CFA Institute

Abstract

Often, the lack of mark-to-market data lures investors into the misconception that alternative asset classes and strategies represent somewhat of a “free lunch.” This article proposes solutions to measuring mark-to-market risk in alternative and illiquid investments. The authors describe how to estimate risk factor exposures when the available asset return series may be smoothed (owing to the difficulty of obtaining market-based valuations). They show that alternative investments are exposed to many of the same risk factors that drive stock and bond returns.

Asset Allocation: Risk Models For Alternative Investments – Introduction

Investors have long recognized that asset-class returns are driven by a common set of risk factors. Asset allocators often use the risk factor approach to improve portfolio diversification and to translate macroeconomic views into expected asset returns. In practice, implementing a risk factor approach to asset allocation requires mapping asset classes to their underlying factor exposures, which can be challenging, especially for asset classes for which the available historical data are limited or biased.

In this article, we propose solutions to measuring mark-to-market risk in alternative and illiquid investments. We describe how to estimate risk factor exposures when the available asset return series may be smoothed (owing to the difficulty of obtaining market-based valuations). We show that alternative investments are exposed to many of the same risk factors that drive stock and bond returns.

Our approach has profound implications for risk estimation in an asset allocation context. For example, Figure 1 shows the difference between adjusted and reported (from index returns) volatilities for several alternative investments, as well as for public markets (equities and bonds). It also shows a measure of autocorrelation, which highlights how return smoothing contributes to the misestimation of volatility. The bottom line is that alternative investments are much more volatile-on a mark-to-market basis—than their reported index returns would suggest. This bias tends to be more pronounced for indices that are smoothed.

In this article, we describe the methodology used to arrive at these adjustments and include other key risk measures relevant to asset allocation. We recognize that there already is a significant body of literature that attempts to estimate risk factor exposures for various individual alternative investments and strategies. However, little research has been done to estimate the risk factor exposures across all alternatives within an internally consistent, unified risk factor framework. Given increased allocations to alternative investments in institutional investors’ portfolios, we see an urgent need to develop a consistent approach that directly integrates the risks of alternative assets with the rest of investors’ portfolios.

Measuring Risk across Alternative Investments

We classify alternative investments broadly into three groups:

  1. private equity and venture capital;
  2. real assets—that is, real estate, infrastructure, farmland, timberland, and natural resources; and
  3. hedge funds and exotic beta strategies (momentum, carry, value, volatility, etc.).

Often, the lack of mark-to-market data lures investors into the misconception that these asset classes and strategies represent somewhat of a “free lunch.” Their relatively high returns appear to come with low risk and significant diversification with respect to traditional asset classes in normal times. This misconception arises because return indices for privately held assets often are artificially smoothed, which biases both volatility and correlation estimates downward.

To address this problem, risk models for private asset classes should rely on public proxies or publicly traded equivalents. Also, the statistical methods used to estimate correlation and volatilities for these assets must be adjusted to reflect the nature of the reporting biases in the illiquid return series.1 To do so, investors must identify the systematic return drivers that affect each of their alternative investments. If the risk model fails to capture the systematic risk factor exposures, diversification benefits may be overestimated.

See full article here by CFA Institute

Leave a Comment