Limiting market risk in financial portfolios

¶ … managing market risk. There are some fundamental differences between market risk and firm-specific risk, although some of the underlying principles of managing this risk are the same. The most common measure of market risk is the value at risk (VaR) calculation, which measures risk in terms of the value of the firm that is at risk during a given point in time, to a specific confidence interval. Outside of that interval, however, incidents can take place that can result in a destruction of firm value greater than the VaR. Examples such as the Amaranth natural gas trading collapse illustrate the importance of a multifaceted strategy for managing market risk. What the Amaranth case shows us is the importance of understanding all of the aspects of market risk that the firm faces, including liquidity risk in big positions. For liquidity risk, there is no equivalent accepted measure to VaR, and this represents challenges to portfolio managers seeking to control their market risk.

This paper analyzes the issue of market risk, some of the measures that can be used to evaluate that risk, and some of the strategies that can be used to manage market risk. Market risk management is an emerging field, and unlike with firm-specific risk there is no one specific objective for portfolio managers with respect to its management. The focus of those studying market risk management, therefore, is to understand the tools that are utilized in measuring and managing market risk. This paper will focus on explaining the underlying concept of market risk and the different tools and theories that can be used to both manage and control market risk. Particular attention will be paid to the case of Amaranth, which illustrates among other things the dangers associated with seeking to eliminate market risk by engaging in strategies that lack diversification.

Introduction

Modern portfolio theory is built around the idea of eliminating firm-specific risk from the portfolio. Every firm comes with risk, but through diversification that risk can be reduced or eliminated, leaving only the risk that is inherent in the market. Risk managers today, having accepted that they can largely eliminate firm-specific risk, have turned their attention to the elimination of market risk. Hedge funds, for example, are designed to have negative betas, enabling them to be used by portfolio managers to help offset market risk. The recent global financial crisis has cast further attention to the issue of market risk, as investors have found it increasingly difficult to find firms and investments that will grow in a slumping market. Broad-based declines in consumer spending depress so many segments of the economy that escaping these declines seems impossible. That is, however, the objective of those seeking to minimize market risk. In a global economy, where portfolio managers have the opportunity to invest anywhere in the world, in any type of product, the opportunity to eliminate a significant portion of a portfolio's market risk should exist. This paper will analyze some of the underlying theories about market risk, and will outline and evaluate some of the different strategies that portfolio managers and finance theorists are undertaking in order to reduce the level of market risk in portfolios.

Summary of Issue

A common measure of risk in a given portfolio is known as "value at risk." This is typically defined as the "maximum dollar portfolio amount that can be lost in a given period of time with a specified level of confidence" (Russon, 2008). The normal confidence level is 5%, and the VaR can be modeled in one of three ways. Measuring risk is the first step to addressing risk. Under normal circumstances, the 95% of the time, market events can be addressed because the portfolio will be designed to eliminate normal risk. It is the remaining 5% of the time that is of the most concern for portfolio managers.

Firm-specific risk is often defined in terms of the company's beta, which reflects the responsiveness of asset price vs. A representative index, typically a stock price vs. A broad stock market index (Jarvela, Kozyra & Potter, 2009). The objective of diversification is to reduce all firm-specific risk, which should deliver to the portfolio a beta of 1.0, the same as the market. Ideally, portfolio managers are seeking to enjoy gains above those of the market, but with downside risk equivalent to the market. In a fully diversified portfolio, the market risk is all that remains, but there remains the 5% of the time that estimates of portfolio reaction to market changes will be off. It is at these times when the portfolio's value is most at risk.

Market risk itself derives from a wide range of elements within the global capitalist economic system. Sullivan (2008) points out that the "capital asset pricing model suggests that investors demand a risk premium in exchange for taking certain investment risks." Part of this compensation reflects firm-specific risk, but the other part reflects systemic risk. There is a disconnect, however, between corporate earnings and the movements of the capital markets. Sullivan argues that this disconnect is reflective of a mispricing of risk in the capital markets. This mispricing presents a challenge for portfolio managers, because it implies that at some point there will be a correction. When risk is underpriced, this leads to asset bubbles which in turn will eventually burst. The result of such bursts is a collapse of the market. In principle, to insult a portfolio from this market risk is to eliminate all such pricing distortions, on both the upside and the downside. The portfolio's value would then not be subject to the seemingly irrational fluctuations of the capital markets. Indeed, Sullivan argues that while the global economy has seen a reduction in risk over the course of the last fifty years, capital markets have continued to have high levels of volatility. This is largely because the providers of liquidity capital have an insatiable appetite for returns that causes them to ignore the risks associated with the returns. The investors, then, are irrational and this causes distortions and volatility in the market (ibid).

When firms fail to pay sufficiently close attention to risk, they place their enterprise at risk. The use of measures such as VaR is intended as the first step in avoiding catastrophe at the hands of distorted capital markets and excessive risk-taking. Chincarini (2008) outlines the case of Amaranth Advisors LLC, a hedge fund that was heavily invested in natural gas futures. The firm was subject to considerable market risk in natural gas futures, and when those futures tanked, it brought the company down. This was the 5% catastrophe in action. When investigators deconstructed Amaranth's plays, it was found that the trading environment in natural gas futures in September 2006 was significantly different from previous years, evidence of a 5% catastrophe. Amaranth was not subject to stop limits and concentration limits, so bore the risk of being heavily exposed to positions. The company lost over $4 billion on its natural gas positions from August 30th, 2006 to September 20th of that year. Contributing to the losses is the lack of downside risk that Amaranth's energy traders had, which caused them to essentially "double down" on their positions, increasing their bets knowing that the only change to recover was if the market recovered and they held deeper positions (Chincarini, 2008).

When Amaranth's market risk was calculated for August 30th, 2006, the VaR was $1.33 billion based on the leveraged positions the company held. The company would have lost nearly $3.3 billion had these positions been held. This was four and a half times the maximum loss that would have occurred any other September with these same positions, so it is evident that September 2006 was subject to unusual market movements that caught Amaranth off guard.

The case of Amaranth illustrates that the first step in addressing market risk is understanding the degree to which the firm or portfolio is exposed to market risk. The strategy that Amaranth was pursuing was calculated to be in pursuit of a profit of almost $400 million, but this came at the risk of losing over $2 billion should an unusually bad market occur.

The company's losses were unusual in part because it had failed to understand the market risk to which it was subject. At these high levels of investment, however, liquidity risk was also a factor. With such strong positions, Amaranth was subject to liquidity constraints that accentuated the losses it faced.

One of the interesting areas of study with respect to market risk is highlighted by the Amaranth case. Hedge funds are typically focused on eliminating market risk, but this typically comes at the cost of focused plays that lack diversification. Thus, the firm is swapping a reduction in market risk for an uptick in investment-specific risk. Such strategies also maintain a significant degree of market risk, in particular the market for a specific commodity with which the firm's asset value is going to be highly correlated. For portfolio managers, the optimal condition is to manage market risk while maintaining the total elimination of firm-specific or commodity-specific risk.

Critical Evaluation

Chincarini's analysis of the Amaranth collapse contains five key lessons for portfolio managers and regulators with respect to managing market risk. Of the ones applicable to firms, the first is that liquidity risk must be accounted for. Liquidity risk in Amaranth's case compounded an already-bad situation. The second recommendation is that liquidity risk measures should be developed that are common -- the way that VaR is common -- so that communication of liquidity risk is consistent throughout the firm and throughout the industry.

The third is that internal risk management practices must be upheld. The best strategies for managing market risk are irrelevant if the firm is not in a position to implement them. In the case of Amaranth, the energy traders were in Calgary while the risk managers were in New York. This made it more difficult to control the firm's exposure in the natural gas futures market. As well internal incentives should reflect the firm's outlook with respect to risk management. Often, successful trading strategies are handsomely rewarded, while those same players are subject to limited downside risk.

When the nature of capital market distortions is analyzed, irrational behavior appears to be present as evidenced by the disconnect between real world market performance and capital market volatility. This irrational behavior may result from situations such as that at Amaranth, where the risk-return tradeoff for the energy traders was not in line with the actual risk-return tradeoff in the market. That the traders increased their bets in the face of steep losses is evidence that such distortions can not only increase the disconnect between the markets and reality but can severely distort risk and return for the company. If market risk is successfully managed, the firm should not be subject to these distortions, so the incentive programs should reflect rational expectations of risk and return on an enterprise-wide level. The last lesson is that spread positions are not arbitrage positions, and must therefore by factored into any evaluation of market risk.

Of these insights, the only one that addresses the issue of market risk on a systemic level is the call for a common measure of liquidity risk. The remainder are largely focused on managing risk at the firm-specific level -- other companies did not suffer the same fate as Amaranth because they did not have the same trading strategies and risk management style. Some of those same strategies were designed to address market risk, at least under normal circumstances, but were misused by Amaranth. As a hedge fun, Amaranth was not expressly intending to eliminate market risk, however. But hedge funds do provide an opportunity to study the issue of market risk further, because many are designed to perform with negative betas, acting as a counter to market movements.

What the Amaranth case highlights is that the same strategies utilized to counter market risk can increase a portfolio's level of firm-specific risk. Amaranth's energy desk was focused on specific natural gas futures plays, a portfolio that can be described as decidedly lacking in diversification. Yet, such a portfolio could have a role to play in countering market risk. If the concept of diversification is applied beyond the stock markets, greater diversification of investment types and investment strategies could result in a portfolio being hedged against stock market risk, if not total market risk. Commodities, currencies, real estate and other tangible goods, emerging markets, fixed income, hedge funds and other investments when put together can reduce a portfolio's total volatility vs. The market, especially when it is considered that some elements often move opposite the market.

Blankfein (no date) argues this point, claiming that market risk is subject to the same principles of quantification as other forms of risk. He outlines a number of different models that can be used to understand market risk. The exponentially weighted moving average model is useful because it measures volatility against the total market while taking into account the reality that high volatility days occur in clusters. This helps account for those 5% incidents with abnormal returns, the same 5% events that threaten firms that lack adequate insulation from market risk. Building on this model is the GARCH3 model, which assumes that variance rates are mean reverting, trending towards a long-run average. These models are used to help firms understand the rate of market risk to which they are subject.

Black-Scholes has also proven applicable to the analysis of market risk. At the heart of Black-Scholes is the concept of a risk-neutral valuation. For derivatives, of course, risk is associated with both expectations of future movements and with time. The risk neutral price of a derivative, as derived from Black-Scholes, is the price at which that derivative should be traded. When applied to the concept of market risk, the risk neutral portfolio is that to which the firm or portfolio manager should strive. The estimates of market expectations are difficult to discern when managing a portfolio consisting of a wide range of investments, but this underlying principle would allow the firm to understand its risk position.

Using Black-Scholes is complex, and may not be practical in all circumstances. Firms therefore gravitate towards simpler measures such as VaR. The VaR can be calculated in one of four ways, but all should deliver the same result (Russon, 2008). A VaR of zero would be perfectly correlated with a lack of market risk altogether, but this is not the objective of portfolio managers; a lack of market risk is a real return-protected Treasury bill. This highlights a key practical difference between market risk and firm-specific risk. While portfolio managers often attempt to eliminate all firm-specific risk from their portfolios, they only seek to manage market risk.