Clients sometimes ask us why we conduct our optimization analyses with discounted dollars (future dollars converted to present-value equivalents), rather than with the actual cash flow of Social Security benefits. After all, they say: “actual cash flows are easy to understand, but ‘discounted’ future dollars is a confusing concept.”
Here is why we discount future dollars: No one believes that $100 to be received 30 years in the future is worth as much as $100 to be received immediately. Our approach is consistent with this universal belief and the behavior that stems from it.
We translate future dollars into present value equivalents, using a 3% “real” discount rate (that is, over and above any inflation). This rate implies, for example, that $100 to be received one-year from now is worth about $97 today. And, a $100 to be received 25 years from now is worth about $49 today. For more discussion on the mechanics of discounting, go here.
Some Social Security optimization programs do not discount; that is, they base their analyses on undiscounted cash flows, treating $100 30-years from now the same as $100 today. So, their approach is totally inconsistent with how actual people think and behave. And, the advice coming from these software packages misleads people into thinking that they should claim benefits later than makes sense from a financial perspective. Of course, many people want to claim sooner rather than later, so giving them misleading advice to claim later than makes sense just causes them to ignore the advice that they just paid for.
Here is an illustration of how the undiscounted cash flow approach seriously tilts the advice toward late claiming. Assume that Andy is single. He is now 60 years old. His SS benefit amount at full retirement is $2,000. His life expectancy is approximately 79.
The naive cash flow analysis suggests that Andy should claim benefits at age 67. In sharp contrast, when we convert future values into present value equivalents with a 3% discount rate, we find that the optimal claiming age for Andy is 62! Discounting really matters, as this example illustrates.
Here is another illustration, for a married couple this time. Let’s use Fred and Ann for our example. Fred was born in 1950, Ann in 1953. His full retirement age benefit is $2,500; hers is $1,500. They both have normal expected life spans: 82 for him and 86 for her.
Using the naive cash flow analysis (no discounting*), the optimal strategy that maximizes that cash flow is:
- Fred files and suspends at 66 in 2016
- Ann claims spousal at 66 in 2018 ($1,250/mo)
- Fred claims retirement at 70 in 2020 ($3,300/mo)
- Ann drops spousal and claims retirement at 70 in 2022 ($1,980/mo)
- Ann switches to survivor’s benefits in 2033 ($3,300/mo).
By the time Ann dies at 86, this strategy will have yielded $1,090,000 in Social Security benefits. (If this time stream of benefits is discounted at 3%, we get a measure of Social Security Wealth equal to $677,000.)
Our Approach: Discounting at 3 Percent
When we discount future dollars using a 3% discount rate**, we would recommend the following for this couple:
- Ann claims retirement at 62 in 2014 ($1,125/mo)
- Fred claims spousal at 66 in 2016 ($750/mo)
- Fred claims retirement at 70 in 2020 ($3,300/mo)
- Ann claims survivor’s benefits in 2033 ($3,300/mo)
By the time Ann dies, this strategy will have yielded $1,071,000 in Social Security benefits, measured on a cash-flow basis. (Converted to a discounted present value of Social Security Wealth, these recommendations produce $685,000–more than the no-discounting recommendations previously discussed).
Notice that by discounting at 3%, we place more emphasis on near-term benefits, which yields recommendations for the couple to start claiming four years earlier than does the naive cash flow analysis.
The naive cash flow recommendations offer a $19,000 advantage over the 24-year time period considered here. But, that approach takes 16 years to gain an advantage over our recommendations, based on discounting at 3%. This point is illustrated in the following graph.
The horizontal axis shows the number of years used in the analysis, from 1 to 25. The vertical axis shows the cumulative losses or gains from following the naive cash flow recommendations versus our recommendations.
By the time the naive cash flow recommendations kick-in (year 5 in the graph), they are already $89,000 behind our recommended approach. And they don’t gain the advantage for another 11 years (year 17 in the graph). And then the advantage is relatively modest ($19,000 received when Ann is about 78 and Fred is about 81, very near the end of his expected life span).
So, by emphasizing near-term benefits over distant benefits, our approach leads to optimal recommendations that typically bring Social Security claimants money years earlier than the naive cash flow recommendations. Moreover, our approach is consistent with how people actually think and behave.
*In this example, discounting future benefits at any rate below 1.75% leads to the same set of recommendations: claim as late as possible.
**In this example, any discount rate between 2% and 5.5% produces the same recommendations, so there is nothing critical about our 3% discount rate.