Model specification of conditional jump intensity: Evidence from S&P 500 returns and option prices

摘要

This study investigates the model specification of the conditional jump intensity under option pricing models having a generalized autoregressive conditional heteroskedastic with jumps (GARCH-jump). We compare three GARCH jumpmodels of Chang, Chang, Cheng, Peng, and Tseng (2018) to examine whether specifying asymmetric jumps in conditional jump intensity can improve the empirical performance. The empirical results from S&P 500 returns and options show that specifying the asymmetric jumps into the conditional jump intensity does improve the in-sample pricing errors and implied volatility errors. However, the out-of-sample results depend on the error measurement.