Assessing the impact of volatility, asymmetry, and directional effects of holidays on Nigerian and Chinese stock exchange price returns using Prophet and APARCH models under variant distributions of innovations

Abstract

Stock price returns are the changes made in the price of a given stock over a specific period and can be expressed as a percentage or price change. A negative return indicates a loss, while a positive return indicates a profit. Unarguably, the research on holidays/events (which may cause sudden fluctuations in stock market prices) must be examined with careful analysis. As such, this paper aims to assess the impact of volatility, asymmetry, and directional effects of holidays/events on Nigerian and Chinese stock price returns using APARCH and Prophet models under normal, skewed Student's t, and Generalized Error Distributions of innovations. Based on the minimum criterion: RMSE, the results revealed that the APARCH model outperformed the Prophet model (under NSE data); while under CSE data, the Prophet model outperformed the APARCH model. More so, there exists a high persistence of the APARCH model. Indicating the impact of past shocks on the current volatility diminishes over time. The Prophet model shows that holidays/events directional effects on Mondays, Tuesdays, and Wednesdays are positive, while for Thursdays and Fridays, the directional effects are negative for NSE data; while for the CSE data, the directional effects of all the holidays from Mondays down to Fridays are positive. Furthermore, there exists a leverage effect (which is negative), implying negative shocks have more impact than the positive shocks under CSE data; while for NSE data, the leverage effect is positive and statistically significant at 5% level under skewed students t while for the normal and GED, the said leverage effect is negative and positive respectively.