Page 129 - 34-1

NTU Management Review Vol. 34 No. 1 Apr. 2024

it is important that they make ample provision for loss reserves.
The coefficients for the group affiliation variables in columns 5 and 6 are positive (i.e.,
0.026 and 0.002, respectively) and significant at the 1 percent level. Thus, insurers that are
members of a group tend to be more over-reserved (less under-reserved) than for single,

stand-alone insurers. Otherwise, the coefficients of the remaining variables are mixed
with respect to signs and significances in the weak and healthy insurer samples. The latter
differences are not unexpected, if it is not appropriate to pool healthy and weak insurers

together in the same regression sample.

5.4 Synthesis of Results
The variables to test the hypotheses are indicator variables, and netting of the
coefficients of these variables allows us to estimate overall effects for publicly-traded

stock insurers that use in-house actuaries. That is, for publicly-traded stock insurers
the overall effect of SOX and the use of an in-house actuary to certify reserves can be
found by netting together the coefficients for the in-house actuary variable and all of its
interaction terms. The net effect for weak publicly-traded insurers associated with the in-

house actuary variable and its interaction terms is 0.005 (i.e., -0.030 + 0.035 + 0 from
column 5 in Table 4). Thus, when all effects are considered, weak, publicly-traded stock
insurers using an in-house actuary after SOX are over-reserved (or less under-reserved);
the corresponding monetary amount is $3.2 million ($631 million in mean total admitted

assets multiplied by 0.005) or 0.5% of total admitted assets at the mean. The amount $3.2
million represents 0.9% of loss reserves and 1.2% of equity.
For healthy insurers, the effect of netting the coefficients is 0.020 (i.e., -0.003 + 0
+ 0.023). Thus, when all effects are considered, healthy publicly-traded stock insurers

using in-house actuaries are over-reserved (less under-reserved) by 2.0% of total admitted
assets or $19.5 million ($976.6 million in mean admitted assets multiplied by 0.020) at
the mean of the sample. As a percent of loss reserves, $19.5 million represents 4.1%, and
this amount represents 3.0% of equity. For nonpublicly-traded stock insurers (i.e., mutual

insurers and privately-held stock insurers), the estimated effect of using in-house actuaries
is merely the coefficient for the In-House Actuary Indicator, which is -0.030 for weak
insurers and -0.003 for healthy insurers in columns 5 and 6, respectively. That is, both
weak and healthy insurers are under-reserved (or less over-reserved), with healthy insurers

121