Sivaprasad G.

Type of the article: Research Article

Abstract
This study empirically examines the dynamic and regime-dependent asymmetric effects of Artificial Intelligence (AI) and Renewable Energy (RE) equity returns on the performance of the U.S. oil sector. Utilizing daily closing prices from December 1, 2017 to October 31, 2025, we employ regime-dependent Vector Autoregression (VAR) models, supplemented by an illustrative Bayesian Time-Varying Parameter (TVP) VAR, to capture heterogeneous transmission mechanisms across bullish and bearish market regimes. The empirical findings show that the individual lagged effects of RE returns on oil returns emerge earlier during bearish markets – a pattern consistent with loss aversion and defensive sectoral rebalancing – whereas the cumulative contribution of RE shocks to oil return variation is larger during bullish phases. In contrast, the relationship between AI and oil shifts direction depending on the market: it is driven by growth when markets are up. Still, it acts as a direct risk transmitter during downturns. Notably, AI and oil equities co-move positively in both regimes, suggesting that their potential diversification benefits as hedges against one another may be limited, although hedge effectiveness is not formally tested here. Collectively, our results underscore that effective investment strategies and policy interventions must be explicitly conditioned on the prevailing market state. Future research avenues are suggested, including subsector-specific analysis and the use of direct technological adoption metrics, such as patent data.