Empirical studies have shown that a large number of financial assets returns exhibit fat tails (leptokurtosis) and are often characterized by volatility clustering and asymmetry. This paper considers the ability of the asymmetric GARCHtype models to capture the stylized features of volatility in USD/KZT exchange rate returns. Therefore, the half-life parameter of the USD/KZT returns series were calculated for three sub-periods. The results revealed that the half-life was 6 days, 16 days and 12 days for 1st sub-period, 2nd sub-period and 3rd sub-period respectively. According to the results, in the presence of asymmetric responses to innovations in the Kazakhstan foreign exchange market, the EGARCH (1.1)-GED model which accommodates the kurtosis of financial time series is preferred. Also, these results show that the USD/KZT exchange rate returns have strong mean reversion and short half-life.
EGARCH, GJRGARCH, APGARCH, USD/KZT exchange rate, Half-life volatility