Embedded value actually has more than one elements - namely the shareholder's share of net assets (surplus or excess assets), VIF (present value of the company's in-force business), and arguably the cost of capital required in terms of the capital requirement to be held in the future years.
If we just look at the VIF element, this is the profit arising from the difference between the best-estimate assumptions (used to project the cash flows) and statutory reserving assumptions (used to calculate reserve required at each future point in time at which the cash flows are calculated).
Hence for calculating embedded value, there is actually two sets of actuarial assumptions required. This is because the cash flows need to be projected for years 1, 2, 3, etc on a best-estimate basis, but the reserves required to be held at each years 1, 2, 3, etc are on statutory reserving basis with compulsory and discretionary margins (assuming calculation on annual basis).
