Abstract
We revisit the tax-loss selling hypothesis as an explanation for the January effect. We expand on prior empirical evidence from municipal bond closed-end funds (CEFs) by extending the sample period by 19 years and adding exchange-traded funds (ETFs). Our sample covers the introduction and rapid growth of municipal bond ETFs, significant changes to municipal bond market structure, and the modernization of tax-loss selling practices. The January effect in municipal bond CEFs has become stronger in recent years and is consistent with the tax-loss hypothesis. The January effect in municipal bond ETFs is smaller and cannot be explained by tax-loss selling.