The costly external financing assumption in capital shock theories of insurance cycles are often attributed to Myers and Majluf (1984). The purpose of this paper is to revisit the Myers and Majluf model and to propose a modified model that better fits capital shock theories. By so doing, this paper attempts to provide justification for existing empirical papers on insurance cycles. Contrary to conventional belief, we argue that while the insight of Myers and Majluf is applicable, their model itself is not appropriate to justify the capital shock theories, since (i) it does not justify price increases at the time of information symmetry; (ii) given information asymmetry, its results imply that an insurer that is severely affected by a shock can always raise capital; and (iii) it does not consider liability that is a main concern of insurers. We consider a modified version of the Myers and Majluf model, by introducing liability. When a shock is small, the results are similar to Myers and Majluf. However, when a shock is large, a pooling equilibrium always exists, in which all insurers cannot raise capital. Interpreting no financing as price increases, this result justifies the argument that costly external financing leads to price increases.