حرکت نزولی، مدیریت ریسک و پرتفوی های بهینه ی مبتنی بر VAR برای فلزات گرانبها، نفت و سهام
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|781||2012||17 صفحه PDF||سفارش دهید||1 کلمه|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : The North American Journal of Economics and Finance, Available online 17 July 2012
Value-at-Risk (VaR) is used to analyze the market downside risk associated with investments in six key individual assets including four precious metals, oil and the S&P 500 index, and three diversified portfolios. Using combinations of these assets, three optimal portfolios and their efficient frontiers within a VaR framework are constructed and the returns and downside risks for these portfolios are also analyzed. One-day-ahead VaR forecasts are computed with nine risk models including calibrated RiskMetrics, asymmetric GARCH type models, the filtered Historical Simulation approach, methodologies from statistics of extremes and a risk management strategy involving combinations of models. These risk models are evaluated and compared based on the unconditional coverage, independence and conditional coverage criteria. The economic importance of the results is also highlighted by assessing the daily capital charges under the Basel Accord rule. The best approaches for estimating the VaR for the individual assets under study and for the three VaR-based optimal portfolios and efficient frontiers are discussed. The VaR-based performance measure ranks the most diversified optimal portfolio (Portfolio #2) as the most efficient and the pure precious metals (Portfolio #1) as the least efficient.
In this high risk and volatile environment, the time is right to examine the downside risk/return profiles for major commodities and stocks. In particular, the downside risk pertains well to the four major precious metals – gold, silver, platinum and palladium – which have risen significantly in terms of global trading and portfolio investments in the recent years, as well as to oil and stocks. The financial and commodity markets had undergone a severe financial crisis in 2007/2008, which turned into a Great Recession, fostering risk aversion and preferences toward safe havens. Despite the ensuing recovery, the mounting risk and uncertainty have confounded investors, portfolio managers and policy-makers. In such an environment, it will be valuable and useful to examine asset behaviors that are not only volatile but also characterized by extreme events like the 2007/2008 financial crisis that affected essentially all asset markets. Standing as hedges and safe havens against risk and during uncertainty, commodities like the precious metals and oil have experienced extraordinary surges in prices and returns in the last few years, which have elevated the potential downside risk and subjected them to black swan-types of events. These assets have therefore become important elements of diversified portfolios. Additionally, stocks have also become very volatile on both sides of the return aisle and had underdone severe extreme events; with high and opposing wild swings being part of their daily trading. Under such circumstances, significant and extreme drops in prices and returns of these assets have become more probable, with potentially damaging consequences on portfolios of individuals and institutions. These circumstances have also made risk management strategies for these high flying commodities and highly volatile stocks more challenging, particularly as the percentages of violations of confidence targets have compounded. The quantification of the potential size of losses and assessing risk levels for individual precious metals, oil, stocks and portfolios composed of them is fundamental in designing prudent risk management and portfolio strategies. Value-at-Risk (VaR) models have become an important instrument within the financial markets for quantifying and assessing market downside risks associated with financial and commodity asset price fluctuations. They determine the maximum expected loss an asset or a portfolio can generate over a certain holding period, with a pre-determined probability value. Thus, a VaR model can be used to evaluate the performance of portfolio managers by providing downside risk quantification, together with asset and portfolio returns. It can also help investors and portfolio managers to determine the most effective risk management strategy for a given situation. Moreover, quantification of the extreme losses in asset markets is important in the current market environment. Extreme value theory (EVT) provides a comprehensive theoretical forum through which statistical models describing extreme scenarios can be developed. There is a cost of inaccurate estimation of the VaR in financial markets which affects efficiency and accuracy of risk assessments. Surprisingly, despite the increasing importance of precious metals and the diversified portfolios that include them as well as other assets and their highly volatile nature, to our knowledge there is only one study that analyzes the VaR for precious metals (Hammoudeh, Malik, & McAleer, 2011), while there are several studies that have worked on oil and stocks’ VaRs. Hammoudeh et al. (2011) concentrate on the four major precious metals only, use relatively older VaR techniques and do not deal with VaR-based optimal portfolio constructions and efficient VaR frontiers. These authors do not distinguish between the risk associated with positive and negative returns which usually display asymmetric behavior. Their study also does not deal directly with volatility clustering. Moreover, it does not include EVT methods which provide quantification of the stochastic behavior of a process at unusually large or small levels. On the contrary, our current study expands the spectrum of asset diversification and deals with events that are more extreme than any others that have been previously observed. Most importantly, it constructs VaR-based optimal portfolios and efficient VaR frontiers of different degrees of diversification and examines their characteristics and performances. It also ranks those optimal portfolios using a VaR-based risk performance measure. The broad objective of this paper is to fill this void in the financial risk management and modern portfolio analysis literature by using more up-to-date techniques and designing optimal diversified portfolios that take into account volatility asymmetry and clustering, with relatively strong emphasis on precious metals which have not been researched adequately despite their potential to provide diversification within broad investment portfolios and hedging capability (Draper, Faff, & Hillier, 2006). To achieve these objectives, the paper computes VaRs for gold, silver, platinum, palladium, oil and the S&P 500 index, using nine estimation methods including RiskMetrics, Duration-based Peak Over Threshold (DPOT), conditional EVT (CEVT), APARCH models (using normal and skewed t-distributions), GARCH-based filtered historical simulation and median strategy. Using different and multiple VaR techniques are of particular importance during high volatility periods like the one the markets experienced during the 2007/2009 Great Recession and its ensuing weak and choppy recovery. The VaR estimates for the different models diverge considerably during these periods, and thus should have pertinent implications for capital charges and profitability. The paper also uses several risk performance evaluations of these techniques including an unconditional coverage test, an independence test and a conditional coverage test. The risk models are also compared under the Basel Accord rules. The optimal VaR-based portfolios and their efficient VaR frontiers are constructed. The portfolio weights suggest that optimal portfolios have more gold than any of the six assets under study. The average portfolio daily returns of the three optimal portfolios differ only slightly. As an annual approximation, we obtain the average returns 9%, 8.625% and 8.5%, for optimal portfolios #1, #3 and #2, respectively. In terms of standard deviation, the most diversified optimal portfolio (#2) has the lowest standard deviation as expected. In terms of statistical properties, the best performers are the conditional EVT and the Median Strategy. Under the Basel II Accord, the performance diverges between the individual assets and optimal portfolios. With individual assets, the RiskMetrics performs poorly and the best performer is the CEVT-sstd model. However, with optimal portfolios the RiskMetrics model is the best performer under the Basel rules, followed by the Median Strategy and the conditional EVT models. In the case of the well-known RiskMetrics model applied to optimal portfolios, there is a discrepancy between the performance using the statistical properties and the performance under the Basel rules. As indicated above, such a study is valuable and useful in light of increases in the weights of commodities, particularly precious metals, in portfolios, especially hedge funds and exchange-traded funds (ETFs). More stringent changes in the Basel accords can have adverse effects on banks, their stocks and the value of their trading portfolios which likely include precious metals and oil, as well as stocks. This paper is organized as follow. After this introduction, Section 2 provides a review of the literature. Section 3 presents the VaR models under comparison. In Section 4 we construct optimal portfolios and their efficient frontiers within a VaR framework. In Section 5, we compare the VaR models using the returns from individual models and form the optimal portfolios constructed in the previous section. Section 6 concludes.
نتیجه گیری انگلیسی
In this paper, Value-at-Risk (VaR) is used to analyze the downside market risk associated with four precious metals, oil and the S&P 500 index. We also construct and rank three VaR-based optimal portfolios and efficient frontiers using these assets. We compute the VaR for the individual precious metals, oil, S&P 500 index and the portfolios, using the calibrated RiskMetrics, the APARCH model, the Filtered Historical Simulation approach, the duration-based POT method, the conditional EVT approach and the Median Strategy. The economic importance of our results is highlighted by calculating the daily capital requirements using the different models. In terms of statistical properties, the best performers are the conditional EVT and the Median Strategy. Under the Basel II Accord, the performance of the different methods in terms of the regulatory capital requirements and days in the red zone diverges between individual assets and optimal portfolios. For individual assets and based on the statistical properties, the RiskMetrics performs poorly while the best performer is the CEVT-sstd model. Based on the average capital requirements and days in the red zone, the performance of RiskMetrics for the individual assets is mixed, giving the lowest average for gold, silver and Brent and the second lowest for the rest of the assets. However, the best performance is still marred with several days in the red zone for silver. Surprisingly, with the three optimal portfolios the RiskMetrics model is the best performer under the Basel rules in terms of both the number of days in the red zone and the average capital requirements. This result has important implications for profitability of the portfolio. The optimal portfolio weights suggest that the three optimal portfolios should have more gold than any of the other assets under study over the sample period. This result contradicts the conventional wisdom which suggests that about 10% of a diversified portfolio should be in gold. The VaR-based performance measure ranks the most diversified optimal portfolio (Portfolio #3 which includes gold, oil and the S&P 500) as the most efficient, and the pure precious metals portfolio (Portfolio #1) as the least efficient. This result underscores the importance of diversifying across different asset classes over diversifying within an asset class even if this class includes a star asset like gold or oil. It has also implications for ETFs which are based on one physical commodity or one asset class. Last but not least, the optimal portfolios give the best performance under the Basel rules for the RiskMetrics model which performs poorly in terms of the statistical properties of individual assets, and thus does not have good reputation.