Thoughts on Extreme Risk in Indonesia

Abstract

Portfolios are commonly optimised using return standard deviation. This article explores a variety of alternate methods for optimising the industry mix of Indonesian portfolios. These include value at risk (VaR) and conditional value at risk (CVaR) using both parametric and nonparametric methods. VaR captures risk at a higher threshold than standard deviation, usually at 95 % or 99 % confidence. CVaR captures extreme risks beyond VaR. The study is unique in its application and comparison of both parametric (normally distributed) and nonparametric VaR and CVaR methods to sectoral portfolio optimization. To capture a range of economic circumstances, the data period incorporates the Global Financial Crisis (GFC) as well as pre-GFC and post-GFC years. The study identifies a fairly narrow band of industries that feature as industries of choice in optimised portfolios across our range of metrics, although their optimal proportions differ across metrics. Using these extreme risk optimizers can assist investors in avoiding high-risk stocks and minimising risk for given return levels. Extreme volatility is also an indicator of underlying problems in an industry, and this comprehensive study on sectoral risk provides important information to lenders, regulators, economic policymakers and governments on the performance and risk of the Indonesian sectoral market and in reviewing sector-based investment and economic policies.