455 Avenue Road, Suite 100, Toronto, Ontario, M4V 2J2
Tel: (416) 966-9675 Fax: (416) 966-9677
The Net Cost of Pure Insurance, or NCPI, is a concept of fundamental importance to advanced life underwriting requiring careful consideration by agents, advisors and consumers. It is understandable that, from an advisor and consumer perspective, the expectation that the NCPI of a policy insuring the same person provided by two different insurers (net amount at risk being equal) should be the same. The reality is that the guidance provided by the Income Tax Act and the Regulations is not complete. In this article, CALU member Florence Marino, B.A., LL.B., TEP, reviews the origins of NCPI, its applications and interpretations.
Table of Contents
The current regime of life insurance policy taxation is largely the result of amendments made to the Income Tax Act in December of 1982. This legislation introduced, among other items:
A policy's ACB is an important concept when dealing with dispositions of the policy. On a policy's disposition, the policyholder receives the ACB (or portion thereof) on a tax-free basis. The reduction of the ACB for the NCPI was a significant development because it reduced the shelter provided by the ACB on a policy's disposition. NCPI was introduced to approximate the pure insurance protection portion of the premiums paid. The reduction in ACB for this amount was made to ensure that tax-deferred growth within the policy is limited to premiums relating to the savings element of an exempt policy.
Regulation 308 provides a description of what insurers must use to determine the NCPI in relation to a policy for purposes of the collateral insurance deduction (subsection 20(1) (e.2)), and the grind to a policy's ACB (paragraph (a) of the description of L in the definition of adjusted cost basis in subsection 148(9) of the Income Tax Act). The Regulation essentially states that the NCPI for a year in respect of a taxpayer's interest in a life insurance policy is the product obtained when the probability of death, computed on the basis of the rates of mortality under a specific mortality table is multiplied by the net amount at risk under the policy. The specific table referred to in the Regulation is the 1969-1975 mortality tables of the Canadian Institute of Actuaries published in Volume XVI of the Proceedings of the Canadian Institute of Actuaries. Interpretation by insurers in the use of this table and the impact on NCPI is discussed in detail below.
The net amount at risk is the amount by which (a) the benefit on death in respect of the taxpayer's interest at the end of the year exceeds (b) the accumulating fund (determined without regard to any policy loan outstanding) in respect of the taxpayer's interest in the policy at the end of the year or the cash surrender value of such interest at the end of the year, depending on the method regularly followed by the life insurer in computing net cost of pure insurance. The impact on NCPI resulting from the insurer's method (i.e. choosing the accumulating fund or the cash surrender value) will also be discussed in detail.
The table referenced in the regulation is a select and ultimate table addressing males and females who are either smokers or non-smokers. What is a select and ultimate table A select rate of mortality takes into account how close a person is to having been underwritten. So, in the year a policy is issued, the table assumes that the person has just been underwritten for insurance and, therefore, mortality would be better than the mortality of the average population. As the person ages and gets further and further from original underwriting, the table reflects worsening mortality so that by the 15th year from the time of underwriting, mortality reflects the mortality of the general population ultimate mortality. After the 15th year, the mortality rates assumed for the particular person continues to reflect ultimate mortality.
In using the table, the insurer would look at the table relating to a person who has the same relevant characteristics as the person whose life is insured. For example, if we are dealing with a female non-smoker age 40, the rate used in this calculation would be the rate in the first column of the table appearing after age 40 on the female non-smoker table. In the following year, when the female is 41 years of age, the rate used would be the second column over in that same row. Comparing this rate to that of a newly underwritten 41-year-old, non-smoking female (i.e., column one of the next row down) shows the difference one year away from underwriting makes in mortality. This difference gets broader and broader as we proceed across the row out to ultimate mortality. Once the end of the row is reached (at age 55 in our example), you have now reached the ultimate portion of the table. One uses this table after age 55 in our example by reading down the last column.
The last row on the table is for age 70. The insurer is instructed to determine NCPI computed on the basis of the specific table referenced in the regulation. Across the industry, the approach to this determination varies when the insured is older than age 70. Some insurers move across the last row to use rates that would be the attained age of the person in question; others utilize ultimate rates using the age of the person in question.
For example, for an 83-year-old female, an insurer may look at a female age 70, move across the row to the 14th column and use that mortality rate. Another insurer may go to the last column of a 68-year-old female and use the ultimate rate for a female age 83. Both of these methods are computed on the basis of the specified table.
As described above, the use of ultimate rates will cause an increase in the assumed probability of death, resulting in a higher NCPI. It is important to understand that although this may be a desirable result to the policyholder where the policy is collaterally assigned because NCPI calculated in this manner would provide a higher deduction (assuming the actual premium is not less than the NCPI as paragraph 20(1)(e.2) limits the deduction to the lesser of actual premium and NCPI), it is an undesirable result if the policyholder is contemplating a surrender or other disposition of the policy since the ACB would be ground down by the NCPI based on the higher ultimate mortality rates. On the other hand, if the policy is corporately owned and intended to be held until death, the ACB grind is a good thing, since the capital dividend account (CDA) credit is the amount of the death benefit received less the corporation's ACB of the policy.
It is clear that NCPI is a double-edged sword. The potential advantage or disadvantage to the policyholder of a particular calculation should have nothing to do with the insurer's determination. Insurers take a position regarding the determination of NCPI computed on the basis of the specified table and would generally consistently apply this methodology across a block of policies, regardless of the policyholder circumstances. So to one policyholder the NCPI determined in a particular way may be beneficial, while to another policyholder this same basis of determination with the same insurer could be detrimental.
There are other areas where insurers must make interpretations to determine NCPI. In making a determination of what mortality rate to use, the insurer is instructed by the Regulation to use the probability of death of a person who has the same relevant characteristics as the person whose life is insured. The tables show males or females who are smokers or non-smokers. What if our 83-year-old female is a substandard risk.
Divergence is found in the industry in determining the NCPI for such a policy. Some insurers may use an age rate-up and the related probability of death of an older aged individual than the insured's actual age. Some insurers use the individual's actual age without any adjustment for the substandard risk associated with the person's state of health. The latter appears to be consistent with the exclusion for substandard risks in the definition of premium found in subsection 148(9).
Another area that requires interpretation is what to do for the NCPI of a joint last-to-die policy. Can a match be found on the specified table for the equivalent single life age (ESLA) relating to the joint last-to-die risk. Is that match a male or a female? For the same two people, each insurer's determination of the ESLA will be different and thus a person who has the same relevant characteristics as the person whose life is insured on the table will be different.
There is an understandable expectation that the NCPI of a policy insuring the same person provided by two different insurers (net amount at risk being equal) should be the same. The reality is that the guidance provided by the Income Tax Act and the Regulations is not a complete code. There are areas that must be interpreted by the insurer in determining NCPI because the legislation does not direct the insurer exactly how to use the specified table in varying situations.
Since an insurer will adopt a methodology across its block of policies, one should not conclude that the insurer adopts a methodology to derive some kind of competitive advantage, because as discussed above, NCPI is a two-edged sword. The advantage or disadvantage of a particular methodology will only play itself out based on the facts and circumstances in which the policy is purchased and ultimately used. This is not in the insurer's control.
From the advisor's perspective, a policyholder may wish to use a policy in a specific way (for example, as a source of funds via partial dispositions of the policy, for deductible collateral insurance purposes and/or to hold until death in a corporation). Understanding the impact NCPI has in the circumstances contemplated is part of the role the advisor assumes in recommending the appropriate product and insurer.
The mortality factor arrived at by using the table is applied to the Net Amount at Risk (NAAR). The insurer determines the NAAR for a policy by taking the death benefit less the accumulating fund (AF) (determined without regard to any policy loan outstanding) or the cash surrender value (CSV), depending on the method regularly followed by the life insurer. So, even if two insurers calculate NCPI using the same mortality factor, the NCPI may still vary based on the method followed in calculating the NAAR. Within the AF method, the outcome may vary between insurers based on the underlying assumptions used to calculate the 1 preliminary term reserve such as the premium payment period, lapses, interest, mortality, etc. Within the CSV method, the outcome may vary between insurers based on items such as surrender charges and bonus assumptions.
The AF method would generally give an equal or lower NAAR (and thus a lower NCPI assuming the same mortality factor is used) to that resulting from the CSV method. As discussed above, a lower NCPI is a benefit if policy values are to be accessed via policy dispositions but a detriment if the intent is to have the death benefit received by a private corporation, or if the intent is to claim the collateral insurance deduction.
Again, since the method chosen by the insurer must be applied consistently, the advantage or disadvantage of a particular methodology will only play itself out based on the facts and circumstances in which the policy is purchased and ultimately used. The method regularly followed in determining the NAAR may well be an area that an advisor may wish to investigate knowing the use for which a client is purchasing a policy.
It should be noted that the Department of Finance has consistently stated that it would like to review policyholder taxation with a view to modernizing the rules. One of the many areas that the Department has mentioned as requiring modernization is the determination of the net cost of pure insurance and the related mortality table referenced in the Regulation. This may be where areas of interpretation in relation to the table are clarified, but it could be a while yet before this occurs.
At present, the areas of variation in interpreting the prescribed table and in determining the NAAR can result in significant differences in the NCPI for a similar policy across insurers. These variations can compound or offset each other. Therefore, advisors should not expect all NCPI calculations to be the same across insurers.
This article was originally published in Volume IV, 2004 of CALU's INFOexchange newsletter. The article was created by CALU Member Florence Marino, who can be reached at firstname.lastname@example.org. The author wishes to thank Greg Cerar, Herb Huck, Steve Krupicz and Russ Lavoie for their comments and contributions to the creation of this article.