OBJECTIVES: National Health Insurance of Taiwan has offered every citizen equal access to health services and financial risk protection from medical expenses. In addition, purchasing private insurances is still popular. The average contracts owned per person is 2.23 in 2012 and has increased to 2.30 in 2013. The premium cost of the National Health Insurance of Taiwan is not a user charged policy, but rather an income-based formula, and most insurances including the National Health Insurance is powered by workforces from the younger cohort. We want to estimate the value of insurance policies by adding the population structure factor. This will reveal the impact to the average value as the population changes. METHODS: We have added the population module into our FinancialCloud system. Data are acquired from Taiwan government open data and dynamic information, i.e., interest rate and foreign exchange rates, are update daily. By computing every possible output of insurance policies for each age cohort (in 5-years interval, e.g. 0-4, 5-10, and so on), we can derive the expected value of a policy. A single year, single term policy pricing may require more than 50 thousand evaluations. For multiple years, terms, and policies, our FinancialCloud is especially suitable for massive, distributed, and parallel computations. RESULTS: Considering a single cohort over the total population can cause price differences of more than 2%, and can even be greater if the term length and population structure changes. For 10 years term, approximately 60% younger cohort is supporting the 40% older cohort, and changes to 68%/32% in 20 years estimate. CONCLUSIONS: For different cohort, the value of a same insurance policy can be different. This factor must be considered to get a better estimate of the policy’s expected value. Rebalancing the premium fee for insurances is necessary if the population structure changes.