We believe that a theory of capital structure based on market timing is themost natural explanation for our results. The theory is simply that capitalstructure evolves as the cumulative outcome of past attempts to time the equitymarket. To our knowledge, this theory of capital structure has not beenarticulated before.There are two versions of equity market timing that lead to similar capitalstructure dynamics. The first is a dynamic form of Myers and Majluf (1984)with rational managers and investors and adverse selection costs that vary acrossfirms or across time. Lucas and McDonald (1990) and Korajczyk, Lucas, andMcDonald (1992) study adverse selection that varies across firms. Choe, Masulis,and Nanda (1993) study adverse selection that varies across time. Consistentwith these stories, Korajczyk et al. (1991) find that firms tend to announceequity issues following releases of information, which may reduce informationasymmetry. Also, Bayless and Chaplinsky (1996) find that equity issues clusteraround periods of somewhat smaller announcement effects. To explain theresults in this paper, these stories require that temporary fluctuations in themarket-to-book ratio measure variations in adverse selection. If the costs ofdeviating from an optimal capital structure are small compared to the resultingvariation in issuing costs, past variation in the market-to-book ratio canthen have long-lasting effects as we observe.