Risk Measures for Alternative-Investment Funds: Part 1

It is generally accepted that alternative-investment funds (alts) are used in part to reduce downside risk. Maximum drawdown, downside deviation, standard deviation, and Sortino and Sharpe ratios are common measures for assessing downside-risk protection. What none of these measures can do is to identify the tail risk (an alts susceptibility to sudden losses) or systemic risk (the potential for co-movements as market prices are falling). In this and some upcoming articles we look at quantifying tail risk and systemic risk for both individual funds and for funds within a classification.

Tail risk is often measured using value at risk (VaR) or conditional value at risk (CVaR, also known as expected shortfall [ES]). Since both of these measures are relatively familiar, we do not directly address them in these articles. Nevertheless, tail risk is a necessary data points for comparing alts within a classification. We write “within a classification” because the drivers of performance and downside-risk reduction vary across classifications but are similar for funds within a classification.

The notion of systemic risk is based on the interdependence of financial institutions (and in our case, alts) with the financial markets in the context of extreme tail events. VaR and CVaR have been criticized for their failure to take into account the escalated risk of spillover effects among financial institutions during episodes of financial crisis. This escalated risk caused by spillover effects is due to the kind and dollar amount of securities the financial institution holds. It has been documented that conditional correlations between assets are much higher in financial market downturns. It has also been shown that the negative feedback of a “loss spiral” or “margin spiral” leads to a joint depression of asset prices. Therefore, systemic risk arises because of the increasing co-movement and linkages among financial institutions’ assets and liabilities during a crisis. In other words, the co-movements between financial institutions and the markets under conditions of distress cannot be measured by their co-movements under normal times. Since systemic risk mainly studies the extent to which extreme values tend to occur together, what really matters for the magnitude of systemic risk is the dependence between the tail risk of individual institutions and the financial system, instead of the characteristics of the marginal distribution for individual firms (i.e., their individual tail risk as measured by VaR or CVaR).

In a significant study by Jiang[1], he writes that it is important that the methodology to assess systemic risk measure the marginal contribution of firm loss to systemic loss, that is, the average return of the firm on the 5% of the worst days, when the market as a whole is in the tail of its distribution[2]. He also writes that measures of ES and a firm’s conditional capital asset pricing model (CAPM) β are highly concordant with each other. We use this relationship between β and the measurement of systemic risk for the alt funds within a classification. In our next article we will outline our method of capturing this β and start to show the results of tail risk and systemic risk computations.