Modeling Value-at-Risk(VaR) in a Small Sized Emerging Financial Market: Evidence from Botswana

Aim of the study: The objective of this study is to model VaR in a small sized rapidly developing financial market in Sub-Saharan Africa which has not only served as a haven for a number of foreign investors, but also has provided the best inflation adjusted returns. This market is of profound interest given that it has received limited attention from policy analysts and previous studies. Methodological framework: This study attempted to employ most of the approaches in modeling VaR, but the results of the diagnostic tests carried out showed that we could only model VaR using either the Basic Historical Simulation (BHS) or the Extreme Value Theory (EVT). Considering the fact that the Peaks over Threshold (POT) is the most preferred choice in academia and industry over the block maxima approach, we opted for the former, which also based on the EVT. The diagnostics were carried out in Eviews, while the parameters of the unconditional EVT and VaR were estimated in Microsoft Excel. Empirical findings: The empirical analysis showed that the tails of the distribution were fatter than in most markets within the emerging market context. These findings do not differ much from previous studies conducted in emerging financial markets. The quantile by quantile plot also showed that the distribution in this market has heavier tails relative to the Student t-distribution. This suggests that any measure of VaR based on assumptions of normality and the Student t-distribution could distort the estimate of Value-at-Risk and have dire consequences on policy decisions. The Kupiec (1995) frequency test showed that both the EVT and BHS cannot be rejected as underlying models to estimate VaR while the Lopez (1998) frequency-of-tail-losses approach which compares and ranks both model showed that the EVT performs better than BHS. Significance: This study bridged the gap in the research literature which has customarily focused on Value-at-Risk measures in “medium and large’’ financial markets in emerging economies by concentrating on a small sized rapidly developing financial market. The findings may also serve as a reference point for most policy makers operating in small sized emerging financial markets