Speculative Volatility and Return Predictability: Evidence from the Japanese Equity Market

Abstract

This study investigates the predictive power of speculative volatility for cross-sectional Japanese equity returns. We construct a latent Speculative Volatility Index (SVI) by applying principal component analysis to the returns of derivative-based and speculative trading indices. The empirical findings indicate that the first two principal components of the SVI significantly forecast stock returns across both market-wide and sector-specific equity indices, exhibiting consistent performance for both in-sample and out-of-sample testing. A long-short trading strategy based on SVI-implied signals from stock-level predictive regressions outperforms an inverse volatility benchmark, thereby validating the SVI’s explanatory power as a novel asset pricing factor.