Risk Management for Event-Driven Funds

Many portfolio strategies are “event driven” (i.e., designed to benefit from price movements caused by corporate events, such as a merger). These strategies involve payoffs with discontinuous and skewed distributions that conventional risk methods do not measure well. This article develops methods to measure the forward-looking risk, based on current positions, of portfolios exposed to such discrete events. The method is applied to independent events and to the more realistic case of events that are not independent. For mergers and acquisitions, empirical estimates of deal-break correlations are positive but low, which implies that most of this event risk is idiosyncratic and diversifiable. The methodology allows assessment of the risk and return of various portfolio structures for event-driven funds.Many portfolio strategies are “event driven” (i.e., are intended to benefit from price movements caused by such corporate events as restructurings, bankruptcies, mergers, acquisitions, or other special situations). Such trading strategies involve payoffs that have discontinuous distributions: Either the event happens or not, which is a binary distribution. In a merger and acquisition (M&A) deal, for example, the price of the target may move up to the offer price if the deal succeeds or may drop sharply otherwise. Such distributions pose a particular challenge to risk management because of their discontinuous nature and the asymmetry of their payoffs. I develop new methods to measure the forward-looking risk, based on current positions, of portfolios exposed to such discrete events.Risk measures for event-driven funds cannot rely on traditional risk models. Such models do not take into account the uncertainty generated by the currently unfolding events; in addition, the history of price movements is not directly relevant for measuring risk. Nevertheless, I show that it is perfectly feasible to construct a forward-looking distribution of portfolio profits and losses based on current positions and to summarize the distribution with the usual value-at-risk measure. To do so, the portfolio manager needs to estimate the probability of success for each deal, the payoffs from success and failure, and the joint correlations across deals.Constructing this distribution reveals a number of interesting insights. For instance, the distribution quantile at a high confidence level measures the “economic capital” required to support the portfolio. It, in turn, defines how much leverage is possible and the expected return on capital.When events are independent, the portfolio return follows a binomial distribution, which is analytically tractable. In this case, the spread of the distribution decreases quickly with the number of deals. Indeed, the economic capital required to support a portfolio of 30 independent deals is six times lower than that required for a single-deal portfolio. Thus, this diversified portfolio could be levered six times more than a single-deal position and still maintain the same level of risk while delivering a higher return on capital.In practice, dependencies exist among events. I present empirical estimates of break probabilities and correlations for a large sample of North American M&A deals over the period 1997 through 2006. The deal-break correlation is estimated to be 0.03, which is positive and significantly so. This correlation can be induced by a dependence on market returns. Indeed, the break probability increases when the stock market falls.With nonzero correlations, the portfolio return distribution can be constructed from the binomial expansion technique or from Monte Carlo simulations. The tails of the distributions are shown to be sensitive to the correlation parameter and, although less so, to the break probability. Positive correlations increase the required economic capital, sometimes substantially. Even so, with a correlation of 0.03, a portfolio of 100 deals can be leveraged three times and still have a level of risk similar to that of a 10-deal portfolio with no leverage. Thus, diversifying among deals reduces risk sharply.Armed with these risk measurement tools, the portfolio manager should search for deals that are profitable yet not too correlated with each other.Note: Pacific Alternative Asset Management Company is a fund-of-hedge-funds investment firm offering strategic alternative investment solutions.