Starting from an environment defined by Rotunno and Emanuel (1987), a hurricane spontaneously forms, reaching an intensity after 10 days (~100 m/s) that is greater than the maximum predicted by Emanuel's theory and those observed in real storms. We propose that this "superintense" storm, which appears similar to those analyzed in several previous studies, is an artifact of a thermodynamic state that is far from radiative--convective equilibrium. The initial storm quickly dissipates and is replaced by a storm that achieves approximate statistical equilibrium after about 20 days. Storm intensity fluctuates in the range of 40--75 m/s for at least 100 days. Periods of sudden intensity change are observed in connection with convectively produced wind-speed maxima at large radii that travel inward to smaller radii, replacing the primary maximum.