The theories posited to explain why a mutual insurer would demutualize can generally be split into two categories: those that suggest mutual insurers converted to the stock organizational form in an attempt to increase firm value and those that suggest the demutualization process allowed the firm's management to extract value for themselves.
We hypothesize mutual insurers had an incentive to demutualize to avoid a unique "surplus tax" imposed by the DEFRA of 1984 when it adopted the "stock model" of taxation for mutual insurers in Section 809.
Thus, some of the insurers remain mutual over the entire sample period, while others demutualize and convert to stock organization form sometime between 1986 and 2004.
16) On one hand, intense competition in urban areas might make it more likely for firms in urban areas to demutualize.
We expect that the estimated coefficient will be positive, consistent with the hypothesis that the IRS equity tax fell most heavily on large equity-based firms and therefore provided these insurers greater incentives to demutualize.
The second specification, Equation (3), allows us to test if the incentive to engage in coordinated risk management changes before and after firms demutualize relative to nonconverting mutuals.
The difference in means is statistically significant for MHC versus nonconverting firms and is consistent with our hypothesis that firms with larger than average equity bases had incentives to demutualize in order to avoid the equity tax.
Firms that fully demutualize have significantly less interest rate risk preconversion than do nonconverting mutuals, and there is a significant decrease in the interest rate risk postconversion.
The average equity base variable is positive and significant, which is consistent with the hypothesis that mutual insurers with larger average equity had greater incentives to demutualize in an effort to avoid losing the deductibility of policyholder dividend distributions.
The only change between the multinomial regression in Table 4 and the logistic regression results shown in Table 3 is our inclusion of a time-varying dummy variable for each state (labeled as MHC State) in an effort to capture the state-specific regulatory environment regarding the decision to demutualize.
Our first observation is that the evidence supporting the access-to-capital hypothesis is stronger for firms that chose to fully demutualize than it is for the MHC conversions.
The estimated marginal effects reported in the last three columns allow us to directly compare the magnitude of the relationship between the exogenous variables in the model and the likelihood that a firm chose to demutualize using a particular method.