In Section 1, we introduced the random future net and gross loss variables. Taking expectations, we have the corresponding policy values,
(~_t V^n =) EPV future benefits – EPV future net premiums
and
(~_t V^g =) EPV future benefits + EPV future expenses – EPV future gross premiums.
We interpret the difference, EPV future gross premiums – EPV future net premiums, to be the expected present value (EPV) of future expense loadings. Thus, we may define the EPV of expenses as
(~_t V^e =~_t V^g – ~_t V^n ) = EPV future expenses – EPV future expense loadings.
If we further assume that (~_0 V^g = ~_0 V^n =0 ) as is customary under the equivalence principle, then (~_0 V^e =0 ). As we have seen in Section 5, it is common for expenses to be high during early policy durations relative to later durations. It is also common for expense loading to be flat, meaning that in general one can anticipate a negative expense value in early durations. A negative expense value is referred to as a deferred acquisition cost, or DAC.
Insurers use policy values for reserves, or liabilities, held against future obligations. At least at early durations, an insurer is required to hold more capital under a net premium basis than a gross premium basis. While it is not common to use gross premium basis in many jurisdictions, in the US the net premium valuation basis has been preferred, in part because it restricts an insurer’s ability to manipulate reserves by making accounting adjustments.