Page 112 - 34-1

In-House Provision of Corporate Services: The Case of Property-Casualty Insurers and In-House Actuarial
Loss Reserve Certification

3.2.2 Specification of the Independent Variables for Hypothesis Testing
We specify an indicator variable (In-House Actuary), which is set equal to one if the
Appointed Actuary is an employee; this variable is set equal to zero otherwise. According
to Hypotheses 1a and 1b, we expect weak insurers to be under-reserved while healthy

insurers to be neither over- nor under-reserved. Therefore, we expect that the coefficient
for the In-House Actuary Indicator will be negative for weak insurers, while the coefficient
for healthy insurers will be insignificant.

Second, an indicator variable denoting Publicly-traded Stock insurers is created.
This variable takes the value of one if the insurer is a publicly-traded stock insurer, and
19
zero otherwise. An interaction variable defined as the interaction of the In-House Actuary
indicator variable and the Publicly-traded Stock indicator variable is created as well.
We expect that weak publicly-traded stock insurers using an in-house actuary are less

under-reserved than weak nonpublic insurers using an in-house actuary under Hypothesis
2. Hence, in the regressions for weak insurers, the coefficient for the in-house actuary
variable interacted with the indicator variable for a publicly-traded stock insurer should be

positive if Hypothesis 2 holds.
To capture the effects of SOX, we create an indicator variable which is set equal to
20
one in 2002 and afterwards, and it is set to zero otherwise. Previous research indicates
that the magnitude of earnings manipulation diminished after the enactment of SOX
in 2002 (e.g., Khurana and Raman, 2004). Hence, we further construct an interaction

variable, In-House Actuary Indicator×Post SOX Indicator×Publicly-traded Stock
Indicator; the Post SOX indicator variable is also interacted with the Publicly-traded Stock
Indicator variable and an External Actuary Indicator variable to create a new variable:

External Actuary×Post SOX Indicator×Publicly-traded Stock Indicator. The External
Actuary Indicator is equal to one if the insurer uses an external actuary and is equal to zero
otherwise.
We expect that after SOX, weak, publicly-traded stock insurers using external
actuaries became less under-reserved under Hypothesis 3a. Therefore, we expect a

19 We discuss other organizational form variables later as they do not relate to any hypotheses.
20 Most studies concerned with the impact of SOX use 2002 as the start of SOX (e.g., Causholli,
Chambers, and Payne, 2014; Iliev, 2010; Knechel and Sharma, 2012; Linck, Netter, and Yang, 2009).

104