This paper aims to identify factors that drive US clean-energy stock price movements in the short and long term, using a wide range of variables representing the carbon emission market, non-green financial assets, non-renewable commodities, macroeconomic fundamentals, investor attention and sentiment, and global stress and uncertainty. The empirical investigation is carried out in the context of the elastic-net regularization (ENET) approach and dynamic simulations of the autoregressive distributed lag (DYNARDL) model, with proper consideration to the potential presence of structural changes. From among 26 candidate variables, the ENET selects the clean technology market, public attention to clean energy, oil, and gold as the primary factors contributing the most to the behavior of clean-energy stock prices. The estimation results of the DYNARDL model suggest that the clean technology market and oil are vital determinants in the short and long run, while public attention and gold tend to affect clean-energy stock prices only in the short run. Furthermore, the respective magnitudes of influence of the four variables are larger in the short term than in the long term. Our findings offer practical implications for socially responsible investors and policymakers.

本文旨在使用代表碳排放市场、非绿色金融资产、不可再生商品、宏观经济基本面、投资者关注度的广泛变量，确定短期和长期推动美国清洁能源股价走势的因素和情绪，以及全球压力和不确定性。实证研究是在弹性网的背景下进行的正则化 (ENET) 方法和自回归分布滞后 (DYNARDL) 模型的动态模拟，并适当考虑结构变化的潜在存在。从 26 个候选变量中，ENET 选择清洁技术市场、公众对清洁能源、石油和黄金的关注作为对清洁能源股票价格行为贡献最大的主要因素。DYNARDL 模型的估计结果表明，清洁技术市场和石油是短期和长期的重要决定因素，而公众关注度和黄金往往只在短期内影响清洁能源股价。此外，四个变量各自的影响幅度短期大于长期。