Quote:
In this approach, spending needs are discounted using the real risk-free rate to match the risk of the cash flows. To be sure, the cash flows are not without risk but their uncertainty is most likely unrelated to market risk factors that would be priced in a normal asset pricing model, making their beta equal to zero. One may argue that, although mortality risk in this context is non-systematic, it is also non-diversifiable. However, mortality risk can be hedged with traditional life insurance allowing the individual to eliminate the non-systematic risk even if it is non-diversifiable. Therefore, discounting spending needs with the risk-free rate is appropriate.
It is tempting to discount spending needs using the expected return of the assets used to fund them. This would be problematic because the risk of the Websters’ spend- ing needs is fundamentally unrelated to the risk of the portfolio used to fund those needs. Merton (2007) draws this distinction in the context of a defined-benefit pension plan. He points out that using the expected return of pension fund assets to discount the liabilities they are intended to fund systematically under-prices those liabilities, and has contributed to the decline of defined-benefit pension plans. Another approach using Monte Carlo simulation with expected returns and volatility is discussed below.
thanks