How volatility model specification affects volatility targeting performance: Evidence from Taiwan

Investors show growing interest in portfolio risk management involving volatility-adjusted allocations. This study develops and assesses a framework for implementing such strategies in Taiwan's equity market, with particular emphasis on how modeling decisions affect investment outcomes. We build on previous research—especially, the conditional approach proposed by Bongaerts et al. (2020) and the ARMA-GARCH framework with jump components discussed by Maheu and McCurdy (2004) and Huang et al. (2024)— by examining the outcomes when different log-return-generating processes are chosen to make predictions. Empirical analysis based on the Taiwan Capitalization Weighted Stock Index shows that this combined specification provides better portfolio performance when it is employed in conjunction with a conditional volatility-targeting allocation rule. A cross-sectional analysis of financial, semiconductor, and food industry indices reveals that conditional volatility targeting strategies have different effects in different sectors. These findings underscore the importance of tailoring volatility models to the underlying distributional characteristics of market log-returns.