Title
Impacts of Time Aggregation Effect of VAR model on Risk Measurements in UK Financial Markets
Abstract
Although many shortcomings of VaR model have been proposed and discussed, VaR is still a commonly used method to estimate potential risk within financial institutions. This paper will focus on the examination of time-aggregation assumption of VaR model in UK financial market. Also, it is necessary to test whether the currently proposed risk forecasting methods such as GARCH method and RiskMetrics approach are realistic in UK market. The results of these examinations will be used to investigate the impact of different methods on the VaR calculation and on the measurement of systematic risk for financial institutions. …show more content…
Based on equation 1.1, 1.2 and the methodology proposed by Baumol, with a certain confidence level c, we can work out the VaR by, rVaR=-W_0 (R^*-μ) Equation 1.3
Rearrange equation 1.3 from the standardized variable we can get rVaR=W_0 ασ Equation 1.3.1
I-2 Research Question
The general idea of VaR model can be shown by equation 1.3.1. However, there is an assumption for financial institutions to process which is known as time aggregation associate with σ in equation 1.3.1. It is related to the risk horizon one would like to generate from a given sample data.
For i.i.d(independent identical distribution) returns, based on the statistical knowledge (Jorion:1997), the measurement to forecast the parameter in equation 1.3.1 is, μ=μ_(unit period)*T Equation 1.4 σ=σ_(unit period)*√T Equation 1.5 and for correlated multi-period returns,
V(∑_(i=1)^T▒R_(t+i) )=σ^2 [T+2(T-1)ρ+2(T-2) ρ^2+⋯+2(1) ρ^(T-1)] Equation 1.6
Where T represents units of period (risk horizon in the calculation) and ρ represents the correlation among the time-series data of returns …show more content…
The problems this article will discuss will be whether this time aggregation under both i.i.d returns and correlated returns is significant in UK market. The discussion will based on the proposed risk forecasting method such as GARCH estimation (Engle:1982), long-horizon GARCH estimation (Christoffersen&Diebold : 2000) and RiskMetrics®. Based on the result of the testing, this article will also focus on figuring out the pattern that the volatility shows in UK financial