How Prevalent is Tax Arbitrage? Evidence from the Market for Municipal Bonds

Although tax arbitrage is central to the literatures on tax capitalization, implicit taxes, and even capital structure, there is little empirical evidence of the extent to which firms actually engage in tax arbitrage. This paper provides some evidence on the topic by focusing on a simple and observable corporate arbitrage strategy in the market for municipal bonds. It poses a puzzle for the literature, however, in that we find little evidence of municipal bond tax arbitrage by non-financial corporations. The overwhelming majority of firms are not engaging in the arbitrage at all and even among those engaged in arbitrage, many firms do less than a safe-harbor amount allowed by the tax authorities. Such a pattern is consistent with the presence of both fixed and marginal (i.e., that depend on size of the position) costs of arbitrage, though we cannot observe what those costs are. The existence of such costs may help explain the municipal bond puzzle and, more broadly, why asset prices often only partially reflect the effects of tax differences. We appreciate helpful comments from John Chalmers (the discussant), Michelle Yetman, an anonymous referee, and workshop participants at Iowa, Ohio State, USC, Wharton, and the 2002 UNC Tax Symposium. Financial support was received from the American Bar Foundation, the Price Waterhouse Foundation, the KPMG Peat Marwick Research Fellowship Fund, University of Chicago and the University of North Carolina.