High-frequency analysis, Model Validation and Intraday risk metrics
https://www.mdpi.com/2227-9091/7/1/10 Background Despite the growing amount of research in the field of high frequency financial data analysis, few studies have focused on model validation and high-frequency risk measures. This study contributes to the literature in the following ways: A rigorous model validation, both in terms of in-sample fit and out-sample performance for the MC-GARCH model under five error distributions is provided. Statistical and graphical tests are conducted to validate the models. One component of the MC-GARCH model is the daily variance forecast. For this purpose, the GARCH(1,1) and EGARCH(1,1) under the five error distributions are compared and the best model among the 10 GARCH models is used to forecast the daily variance. The modelling and forecasting performance of the MC-GARCH model under different distributional assumptions is assessed in this study. The 99% intraday VaR is forecasted and three backtesting procedures are used. This is the first study to assess the VaR predictive ability of the MC-GARCH models by using an asymmetric VaR loss function. This is the first study to forecast the intraday expected shortfall under different distributional assumptions for the MC-GARCH model. Again, three backtests are used including the recently proposed ES regression backtest. Due to the high importance of risk management, the results of this study may contribute in many fields. This study is highly relevant to the banking industry since banks are required to calculate risk metrics on a daily basis for internal control purposes and for determining their capital requirements. Risk measurement is also essential to the insurance industry from the pricing of insurance contracts to determining the Solvency Capital Requirement (SCR) and therefore the results of this study might be useful. Any other organisation having an exposure to some kind of financial risk might benefit from this study.