Loading...

Spillover Effects in Returns and Volatilities to Estimate the Value at Risk of Portfolio, Consist of Gold, Foreign Exchange and Stocks

Moftakhar Daryaie Nejad, Kobra | 2013

477 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 44184 (44)
  4. University: Sharif University of Technology
  5. Department: Management and Economics
  6. Advisor(s): Keshavarz Haddad, Gholamreza
  7. Abstract:
  8. Accurate modelling of volatility (or risk) is important in finance, particularly as it relates to the modelling and forecasting of value-at-risk (VaR) thresholds. As financial applications typically deal with a portfolio of assets and risk, there are several multivariate GARCH models which specify the risk of one asset as depending on its own past as well as the past behaviour of other assets. So we mustn't ignore return and volatility spillover effects between different assets in the portfolio for estimating the portfolio value at risk. Many studies in financial economics in recent decades investigate the spillover effects and modelling it. In this research we analyze the importance of considering spillover effects when forecasting financial volatility among three international markets: Gold, Stocks and Foreign Exchange. We use the logarithmic returns of the assets - ounces of gold, the euro-dollar exchange rate and America stock index S&P500- from the first business day of 2000 untill 9/21/2012 in order to identify the relationship between these three international markets. Identification returns transmission between markets is provided by using Vector AutoRegressive (VAR) model. Volatility spillover effects could be measure by the Multivariate Generalized AutoRegressive Conditional Heteroskedasticity (MGARCH) models. The MGARCH models can capture several stylized facts of financial returns, such as time-varying volatility and clustering of volatility. Therefore we use VAR-MGARCH model for identifying information spillovers between three markets and introduce value at risk in order to measure portfolio risk. We estimate the value at risk of portfolio relying on parametric approach (assuming multivariate normal and t-student distributions) at 99% confidence level for the forecast horizon of one day with two rolling windows that included 500 and 750 observations. After forcasting VaR, we test the validity of our predictions by two-step back-testing approach. In the first step the adequacy of predictions was tested by unconditional and conditional coverage tests.Then the adequacy predictions were being ranked by Lopez loss function, Total accumulated losses and Sener loss function. The empirical results suggest that spillover effects are statistically significant and the VaR threshold forecasts are generally found to be sensitive to the inclusion of spillover effects in any of the multivariate models considered. Ignoring this sensitivity is made to overestimate of the portfolio’s value at risk and, therefore, lead to inefficient allocation of resources to cover the risk of the asset portfolio. Also, by changing the length of the rolling window, the performance evaluation of different methods that estimate value at risk, will change
  9. Keywords:
  10. Volatility ; Value at Risk ; General Autoregressive Conditional Heteroskedastic (GARCH) ; Ranking ; Return Spillover ; Rolling Window

 Digital Object List

 Bookmark

No TOC