Mortality Tables: CSO, VBT, and Projection Scales

Standard Mortality Tables in Life Insurance

Actuaries use several standard mortality tables depending on the application. The Commissioners Standard Ordinary (CSO) table is prescribed for statutory reserve calculations and nonforfeiture values in life insurance. The Valuation Basic Table (VBT) reflects raw mortality experience without the margins added for the CSO and is used for principle-based reserving and other analyses requiring best-estimate mortality. Understanding the differences between these tables, their construction, and their appropriate applications is fundamental to actuarial practice in life insurance.

The current CSO table (the 2017 CSO) replaced the 2001 CSO table and reflects significant improvements in insured-lives mortality over the intervening period. The table provides mortality rates by age, gender, and tobacco use status, with separate rates for smokers and nonsmokers. The CSO includes margins above best-estimate mortality to ensure reserve adequacy, consistent with its regulatory purpose. These margins mean that reserves calculated using the CSO will, on average, exceed the expected present value of future claims, providing a buffer against adverse mortality fluctuations.

The Valuation Basic Table

The VBT is developed from the same experience data as the CSO but without regulatory margins. It represents best-estimate mortality for insured lives and serves as the starting point for mortality assumptions under principle-based reserving (PBR). The 2015 VBT provides rates by age, gender, tobacco status, and underwriting classification (preferred best, preferred, residual standard, and others). The granularity of the VBT allows actuaries to develop mortality assumptions that reflect the specific risk characteristics of their company's insured population.

Under PBR, companies are required to use the VBT as a starting point and adjust for their own experience through credibility-weighted blending. Companies with large, credible experience data sets may develop mortality assumptions that deviate significantly from the VBT, while smaller companies with less credible data rely more heavily on the standard table. The credibility methodology, which determines how much weight to give company experience versus the standard table, is a critical component of the PBR framework.

Pension Mortality Tables

For pension and retirement plan applications, the SOA's Retirement Plans Experience Committee develops mortality tables based on the experience of large defined benefit pension plans. The RP-2014 table (and its successors) provides base mortality rates for four categories of pension plan participants: employees (still working), healthy annuitants (retired and receiving benefits), disabled annuitants, and juvenile dependents. Each category has distinct mortality characteristics reflecting differences in health status, socioeconomic factors, and selection effects.

The Pri-2012 table is a more recent mortality study that provides updated base rates and additional granularity, including separate tables by collar type (white collar, blue collar, and mixed). The distinction between white-collar and blue-collar mortality is significant: blue-collar workers generally experience higher mortality rates at every age, reflecting differences in occupation, income, lifestyle, and access to healthcare. Pension actuaries must select the appropriate table based on the characteristics of their plan's participant population.

Mortality Improvement Scales

Mortality has generally improved over time, and projection scales capture this trend for forward-looking analyses. The SOA publishes mortality improvement scales (MP-2021 and subsequent updates) based on historical mortality data and expert judgment about future trends. These scales project how mortality rates will change in future years and are critical for pension and annuity valuations, where the present value of future benefits increases as people live longer.

The two-dimensional nature of these scales (varying by both age and calendar year) allows for different improvement rates at different ages and for the gradual convergence of near-term and long-term improvement assumptions. Near-term improvement rates are based on recent observed experience and may be higher or lower than the long-term ultimate rate, depending on recent mortality trends. The ultimate improvement rate represents the assumed long-term annual rate of mortality reduction, typically set at 0.5% to 1.5% per year depending on age and gender.

The MP scales use a model that blends recent experience (which captures current trends, including any short-term disruptions) with a long-term assumption. The transition from near-term observed rates to long-term assumed rates occurs over a convergence period, typically 10 to 20 years. This structure means that the projected mortality rates in the near future are heavily influenced by recent experience, while rates projected decades out are driven primarily by the long-term assumption.

Impact on Valuations

The choice of mortality table and improvement scale has a material impact on actuarial valuations. For a pension plan, switching from the RP-2014 table with MP-2014 improvement to a more recent base table with updated improvement can change the present value of benefit obligations by several percent. For a large plan with billions of dollars in liabilities, this translates into hundreds of millions of dollars in valuation changes.

In life insurance, the mortality table affects statutory reserves, nonforfeiture values, and the premium rates charged to policyholders. The transition from the 2001 CSO to the 2017 CSO generally reduced statutory reserves because the newer table reflects lower mortality rates, allowing companies to hold less reserve per dollar of insurance in force. However, the impact varies by age, gender, and product type, requiring careful analysis of each company's in-force block.

Developing Company-Specific Mortality Assumptions

While standard tables provide useful benchmarks, most actuarial analyses require company-specific mortality assumptions. Life insurers analyze their own claims experience to determine how their mortality compares to the standard table, often expressing the difference as an A/E (actual-to-expected) ratio. An A/E ratio below 1.00 indicates better-than-standard mortality (fewer deaths than the table predicts), while a ratio above 1.00 indicates worse-than-standard experience.

The credibility of company-specific experience depends on the volume of data available. Standard credibility theory provides formulas for blending company experience with the standard table based on the number of claims and exposed lives. Companies with large blocks and many years of experience can rely heavily on their own data, while smaller companies must lean more on industry tables. The actuary must also consider whether recent experience reflects sustainable trends or temporary fluctuations, and whether changes in underwriting standards or product mix warrant adjustments to historical experience.

Study Tips and Exam Relevance

Mortality tables and improvement scales are tested extensively on SOA exams, particularly in the Individual Life and Annuities track and the Retirement Benefits track. Candidates should understand the construction and purpose of the CSO, VBT, and RP tables; the structure of mortality improvement scales; and the methods for developing company-specific assumptions through experience studies and credibility weighting. Practice problems involving the calculation of reserves using different mortality tables and the quantification of mortality improvement's impact on annuity values build the skills needed for both exams and professional practice.