Simple Deterministic View of the PV of Spending Rates vs Longevity

I had no real reason for doing this other than I was curious what it looked like plus my SO forced me to think about this after I got a "don't worry, pat on the head" kind of thing.  No one seems to think much about longevity uncertainty these days except the academic researchers.


I realize now after the fact that the spend rate examples are probably a little high for most realistic situations but I didn't want to go back and change it.  Note that the shift from one line in the chart to another is more or less what I feel like I'm being asked personally to consider in extra spending over a long term timeframe in order to accommodate more travel.  The shift, btw, would be as true at 50k in annual spending as it is at 80k or above. Either way I've convinced myself that it would be a material change in spending and risk but I will keep my thoughts to myself for now.

I know it's much cooler to do complex simulations these days but sometimes a simple, deterministic look at this retirement stuff is informative.  Here, then, are two charts that compare the NPV of different spending levels inflated at 3%. For each longevity estimate on the x axis, a PV of spending at each level (to that x-axis age estimate) is calculated on the y axis.  The first chart has a higher discount rate (5%) and no provision for a long term care estimate. The second chart has a lower discount rate (3%) and folds in an estimate for long term care of $250k in NPV terms[1].  The point of view is that of a 58 year old early retiree.  The planning range assumes that the average longevity estimate for a 58 year old (81) is a good estimate for the low end of the range and that if a 58 year old were to survive to age 85, a third quartile longevity estimate for the cohort at that time (94-95) is probably a good conservative estimate for the upper range.  But who knows.  Maybe we will all live to 200.

X axis: Age / Longevity with respect to a 58 year old
Y axis: NPV of various spend rates inflated at x% and discounted at y%
Planning range low: Average longevity for 58yo using Soc Sec tables
Planning range high: 3rd quartile longevity est if one survives to 85
Spend inflation: 3%
Discount: approximation for real rate of return, first 5% chart1 then 3% chart2
Long term care estimate: 0 in first chart, 250k NPV in second chart
Spend step function for 120k: drops 17% at age 69 and 30% at age 86.
  - this approximates Blanchett's estimate of 1.5-2% annualized reductions in spending in later life.

Conclusions:

The point of this exercise was not really about "conclusions" but if forced I would say again: don't retire earlier than necessary and be wary of spending at levels higher than necessary at all points down the line.  But that's just me.  

------------------------------
[1] this estimate is arbitrary but I remember seeing some research somewhere or other that had a swag in this range for LTC in PV terms.  Someday I'll go back and look at it.  Maybe just consider this a reserve for "spending shock."