Electronic copy available at: http://ssrn.com/abstract=1762116 The Relative Leverage Premium ∗ Filippo Ippolito Universitat Pompeu Fabra filippo.ippolito@upf.edu Roberto Steri Bocconi University roberto.steri@unibocconi.it Claudio Tebaldi Bocconi University claudio.tebaldi@unibocconi.it First version: February 14th, 2011 This version: February 28th, 2012 Abstract The existing empirical evidence does not yet provide a clear understanding of how leverage and expected equity returns are related. While in some studies the relationship between financial leverage and returns is positive, in others returns are either insensitive or decrease with leverage. We re-examine this evidence by explicitly accounting for the dynamic nature of a firm’s optimal leverage policy in the presence of frictions. Specifically, we allow firms to temporarily deviate from their optimal capital structure due to adjustment costs. For each firm we estimate target leverage, and compute relative leverage as the difference between observed and target leverage. We find that relative leverage is positively and significantly related to expected equity returns, and has a dominant effect over size and book- to-market. The relative leverage premium shows a remarkable symmetry for over- and under-leveraged firms. JEL Classification: G12, G32 Keywords: leverage, cross section of returns, target leverage, dynamic capital structure, financial frictions * We would like to thank Micheal Brennan, Mark Flannery, Vidhan Goyal, Christopher Hennessy, P´ eter Kondor, Arthur Korteweg, Eric Lam, Juhani Linnainmaa, Michael Mueller, Alessio Saretto, Stephen Schaefer, Lukas Schmid, Pietro Veronesi, and Kuan Xu for valuable comments and discussions, and Samad Javadi for research assistance. We also have benefited from comments from seminar parteci- pants at Universitat Pompeu Fabra, at Goethe University of Frankfurt, at the 18th Annual Conference of the Multinational Finance Society, at the 7th CSEF-IGIER Symposium on Economics and Institu- tions, at the 2011 Northern Finance Association Conference, at the 2012 American Finance Association Annual Meeting, and at the 2012 Midwest Finance Association Annual Meeting. All remaining errors are our own. Filippo Ippolito is associated with the Department of Economics and Business, Universi- tat Pompeu Fabra and with the Barcelona Graduate School of Economics, Ramon Trias Fargas 25-27, 08005 Barcelona, Spain. Roberto Steri and Claudio Tebaldi are affiliated with the Department of Fi- nance at Bocconi University, Via R¨oentgen1, 20136 Milan, Italy. Corresponding author: Roberto Steri, Department of Finance, Bocconi University, Via G. R¨oentgen n. 1 - 20136 - Milano, Room 2-C3-04.