Proceedings 2nd ISI Regional Statistics Conference, 20-24 March 2017, Indonesia (Session IPS21)
Volatility modelling using Box
–
Cox stochastic volatility model:
Evidence from Bank Indonesia (2010
–
2015)
Didit B. Nugroho*
Dept. of Mathematics, Satya Wacana Christian University, Salatiga, Indonesia – [email protected]
Tundjung Mahatma
Dept. of Mathematics, Satya Wacana Christian University, Salatiga, Indonesia – [email protected]
Yulius Pratomo
Dept. of Economics, Satya Wacana Christian University, Salatiga, Indonesia – [email protected]
Abstract
Volatility of the foreign exchange rate in the Indonesia’s central bank is modelled and estimated. The empirical analysis is based on daily data on selling four foreign currencies to Indonesia rupiah (IDR); these include the Swiss franc (CHF), the Euro (EUR), the British pound (GBP), the Japanese yen (JPY), and the US dollar (USD) over the period of January 2010 to December 2015. This study employs a class of nonlinear asymmetric stochastic volatility model by applying the Box–Cox transformation to the volatility equation and calls it “Box–Cox asymmetric stochastic volatility (BCASV) model”. The empirical results show that the CHF and GBP returns series are in favour of the Box–Cox specification in terms of the 95% highest posterior density interval for the Box–Cox parameter. Results also indicate that the leverage effect is absent for all returns series. In particular, we find evidence of significant positive correlation between returns and volatility in adopting the CHF, JPY, and USD returns series. Furthermore, the log-marginal likelihood scores confirm that the BCASV model is superior to the asymmetric stochastic volatility (ASV) model for CHF, EUR, and GBP returns series.
Keywords: