Portfolio Value-at-Risk with Time-Varying Copula: Evidence from Latin America

Model risk in the estimation of value-at-risk is a challenging threat for the success of any financial investments. The degree of the model risk increases when the estimation process is constructed with a portfolio in the emerging markets. The proper model should both provide flexible joint distributions by splitting the marginality from the dependencies among the financial assets within the portfolio and also capture the non-linear behaviours and extremes in the returns arising from the special features of the emerging markets. In this paper, we use time-varying copula to estimate the value-at-risk of the portfolio comprised of the Bovespa and the IPC Mexico in equal and constant weights. The performance comparison of the copula model to the EWMA portfolio model made by the Christoffersen back-test shows that the copula model captures the extremes most successfully. The copula model, by estimating the portfolio value-at-risk with the least violation number in the back-tests, provides the investors to allocate the minimum regulatory capital requirement in accordance with the Basel II Accord.