기대수명 증가와 종신연금

Abstract

As individuals' life expectancy increases, retirees are exposed to greater risk of out-living their wealth. This longevity risk is one of the most serious risks to individuals who prepare for retirement, because when they do not have enough money to fund their retirement years, people will experience deteriorated retirement level of living and the cost for supporting elderly will be a great burden for a society. So far, research and practical focus of retirement planning have been on accumulating enough wealth for retirement, in other words, saving for retirement has been the major retirement planning goal for many people. However, retirement wealth management during retirement is also important as people's retirement years become longer due to the increased life expectancy. The aim of this research is to explore the relationship between increased life expectancy and life annuity effectiveness. Specifically this research tries to provide comprehensive reviews on previous research on annuitization, to set up a model to calculate annuity prices based on increased life expectancy, to explore average number of years of needing extra support in retirement to cover desired level of living, and to calculate average level of living in retirement years. Annuitization refers to converting retirement wealth in life annuity form and produce constant stream of income during retirement. Therefore, researchers have been suggested full annuitization of retirement wealth as a utility maximization strategy under longevity risk(Yarri 1965). However, annuity markets are under-utilized compared to annuity’s known benefits of hedging longevity risk. Many possible reasons of the annuity puzzle have been suggested in literature (e.g., Lockwood 2010; Vidal and Lejárraga 2004). One of the reasons of the annuity puzzle is the low interest rates and actuarially unfair set up of annuity prices due to the adverse selection of annuity products. Therefore, some people will prefer managing their retirement wealth by investing other financial products such as stocks and bonds instead of buying annuity products(i.e., annuitization). Previous literature has been compared the advantages and disadvantages between annuitization and portfolio management of retirement wealth and suggested utility maximization time of annuitization for retirees under various risk tolerance levels (eg., Dus, Maurer and Mitchell 2005; Kingston and Thorp 2005; Milevsky and Young 2007). This research tries to measure annuitization effectiveness on hedging longevity risk by measuring average number of years of needing extra support in retirement to cover desired level of living, and calculating average level of living in retirement years. Using 2011 Korean Survey of Household Finances, the average non-retired 60 year old household wealth was identified to simulate retirement income funding during retirement based on various levels of annuitization. To run the simulation, it was assumed that the household would use its entire wealth except its primary residence on funding retirement income. The retirement wealth was assumed to invested annuity and portfolio and the portfolio was assumed to be divided into 60% stock and 40% bond. The annuitized wealth level was varied 5% increment starting from 0%. Every year retired household would receive annuity income from the annuitization wealth based on its investment size and the shortfalls to their desired retirement income was assumed to be supported from the portfolio. The remaining portfolio asset after the shortfall withdrawal was assumed to be re-invested into stock and bond as 6:4 without any cost. The portfolio asset was assumed to follow lognormal probability distribution. The average rate of return and standard deviation for the past 10 years were used in the simulation. To set up an annuity prices, market mortality tables from the Korean Insurance Development Institute were used. To capture life expectancy changes, the two most recent market mortality tables have been used. The 6th market mortality table was published in 2009 and the 7th market mortality table was released in July 2012. Using Crystal Ball software, Monte Carlo simulation method has been utilized to explore the average number of years of needing extra support in retirement to cover desired level of living and average level of living in retirement years. After 10,000 simulation of each scenario it has been found that increased life expectancy negatively influences on the average number of years of needing extra support in retirement to cover desired level of living and average level of living in retirement years. As life expectancy increases, retirees will need to support longer years during retirement to cover desired consumption level using extra assets and the average level of living during retirement will be deteriorated. When retirees annuitize more portion of their wealth, they will need to support longer years with extra cost but the overall average level of living during retirement will be improved. The findings of this research suggested that retires and practitioners will need to consider their desired level of living and life expectancy in retirement asset management. This research utilized a model with one time decision at retirement therefore, the findings of this research could provide practical example of retirement asset management at retirement, even though phased decision making and gradual annuitization could be better options for some retirees. Readers should keep in mind that the mortality risk fees from insurance companies were not fully incorporated in annuity pricing model therefore; annuity income examples from this research are not fully comparable to commercial life annuity products.