64
sensitivity analysis.
The work described in this chapter built the foundation to test the hypothesis
on whether a hazard function has a significant change-point, which is illustrated in
Chapter 5.
4.1 Basic framework
We want to understand the impact of the GLB Act on insurance companies from two
perspectives, i.e. to the companies’ survival and to the Acquisition among insurance
companies. When we looked at the data, the percentage of acquisitions is relatively
high, which shows that being acquired is an important choice that insurers have when
their survival ability is challenged. Hence, a study of acquisitions is essentially valu-
able for understanding the impact of the GLB Act. Given the respective definitions
of an event and dependent censoring in Section 3.1, we applied the proposed method
to two scenarios.
In the first scenario, the purpose is to identify what factors, if any, have significant
impacts to insurance firms’ survival, denoted as “for bankruptcies”. Here, the event
of interest is “going bankrupt”.
In the second scenario, we focused on the insurance companies that are likely and
willing to be acquired. The purpose is to identify what factors, if any, have signifi-
cant impacts on insurance firms’ acquisition, denoted as “for acquisitions”. Here, the
event of interest is “pursuing acquisitions”. This scenario is not quite the same as