Alexandra Livada

Evaluating value-at-risk models before and after the financial crisis of 2008: International evidence

Purpose - The purpose of this paper is to focus on the performance of three alternative value-at-risk (VaR) models to provide suitable estimates for measuring and forecasting market risk. The data sample consists of five international developed and emerging stock market indices over the time period from 2004 to 2008. The main research question is related to the performance of widely-accepted and simplified approaches to estimate VaR before and after the financial crisis. Design/methodology/approach - VaR is estimated using daily data from the UK (FTSE 100), Germany (DAX30), the USA (SP classic GARCH(1,1) model of conditional variance assuming a conditional normally distributed returns; and asymmetric GARCH with skewed Student- Findings - The paper provides evidence that the tools of quantitative finance may achieve their objective. The results indicate that the widely accepted and simplified ARCH framework seems to provide satisfactory forecasts of VaR, not only for the pre-2008 period of the financial crisis but also for the period of high volatility of stock market returns. Thus, the blame for financial crisis should not be cast upon quantitative techniques, used to measure and forecast market risk, alone. Practical implications - Knowledge of modern risk management techniques is required to resolve the next financial crisis. The next crisis can be avoided only when financial risk managers acquire the necessary quantitative skills to measure uncertainty and understand risk. Originality/value - The main contribution of this paper is that it provides evidence that widely accepted/used methods give reliable VaR estimates and forecasts for periods of financial turbulence (financial crises).

Rolling-sampled parameters of ARCH and Levy-stable models

In this article an asymmetric autoregressive conditional heteroskedasticity (ARCH) model is applied to some well-known financial indices (DAX30, FTSE20, FTSE100 and SP500), using a rolling sample of constant size, in order to investigate whether the values of the estimated parameters of the model change over time. Although, there are changes in the estimated parameters reflecting that structural properties and trading behaviour alter over time, the ARCH model adequately forecasts the one-day-ahead volatility. A simulation study has been carried out to investigate whether the time-variant attitude holds in the case of a generated ARCH data process revealing that either in that case the rolling-sampled parameters are time varying. The rolling analysis is also applied to estimate the parameters of a Levy-stable distribution. The empirical findings support that the stable parameters are also time variant.

Evaluation of realized volatility predictions from models with leptokurtically and asymmetrically distributed forecast errors

Accurate volatility forecasting is a key determinant for portfolio management, risk management and economic policy. The paper provides evidence that the sum of squared standardized forecast errors is a reliable measure for model evaluation when the predicted variable is the intra-day realized volatility. The forecasting evaluation is valid for standardized forecast errors with leptokurtic distribution as well as with leptokurtic and asymmetric distributions. Additionally, the widely applied forecasting evaluation function, the predicted mean-squared error, fails to select the adequate model in the case of models with residuals that are leptokurtically and asymmetrically distributed. Hence, the realized volatility forecasting evaluation should be based on the standardized forecast errors instead of their unstandardized version.