According to the classical theories of commercial annuity demand, households should more or less annualize their wealth to cope with individual longevity risk. However, in reality, household participation in commercial annuity market is limited which gives rise to annuity puzzles. Classical theories generally adopt the life cycle model as the research paradigm, which assumes that households will calculate the optimal path of future consumption and savings with the goal of maximizing lifetime utility. The ensuing question is whether households have set about calculating their needs to prepare for the future, as assumed in theory. In short, have households tried to calculate retirement needs? The answer to this question is seen as an indicator of whether households have retirement planning propensity. The heterogeneity of rational and farsighted optimization will result in the difference of household economic decision-making. Households with retirement planning propensity are more in line with the assumption of life cycle model, that is, they follow intertemporal optimization of economic decision-making. As a result, these households are more likely to obey the classical life cycle model to make relatively rational economic decisions. Existing literature has proved that retirement planning propensity can explain household heterogeneity in savings level, risk asset allocation, wealth accumulation and other aspects, but doesn’t involve in commercial annuity demand. Based on this, this paper empirically studies whether retirement planning propensity will boost household commercial annuity demand. There are two centralcontributions of this paper: First,it explores the role of retirement planning propensity from the perspective of household risk management, which contributes to the existing literature on the effects of retirement planning propensity on household economic outcomes; Second, this paper studies for the first time the relationship between retirement planning propensity and household commercial (annuity) insurance demand, suggesting that retirement planning propensity may be a behavioral factor to help explain limited participation in the commercial annuity market. Specifically, this study utilizes the data of 2014 China Family Panel Studies (CFPS) to conduct an empirical analysis. It uses the question asking whether household calculating retirement needs to proxy for retirement planning propensity. The results of multivariate Probit model show that: (1) Households who have calculated retirement needs will systematically increase the probability of holding commercial annuity, and the effect of retirement planning propensity accounts for more than 40% of the average participation level of the sample; (2) The calculation results of retirement needs show adverse selection. Expected retirement needs may contain information about subjective life expectancy, so the higher the expected retirement needs, the more likely the households are to hold commercial annuity because of more expected benefits; (3) Retirement planning propensity will also increase household demand for other types of commercial insurance besides commercial annuity, thus improving the participation level of household commercial insurance as a whole. It indicates that individual longevity risk is not the only risk faced by households when formulating retirement plan. The above results are robust after considering endogeneity problems such as reverse causality. The policy implications of this paper are as follows: The government should encourage households to use different ways to calculate retirement needs or provide the services needed to calculate retirement needs, reduce the searching and processing cost of information about retirement planning, improve the retirement planning propensity and assist in the formulation of specific plans. This will help promote the development of the commercial annuity market and even the whole commercial insurance market, so as to make it play a greater role in household risk management.