No. of Recommendations: 22
Here is another one: Rich, Broke or Dead? Post-Retirement FIRE Calculator: Visualizing Early Retirement Success and Longevity Risk. Description: "This interactive post-retirement fire calculator and visualization looks at the question of whether your retirement savings can last long enough to support your retirement spending and combines it with average US life expectancy values to get a fuller picture of the likelihood of running out of money before you die."..
That would count as another fine example of the sort of "calculator" that should be completely banned.
The assumptions are stated, but few will read them, and even fewer will realize that the assumptions are absolutely not reliable, so neither are the results.
The biggie can be summed up here: "if your retirement portfolio survives most historical cycles, there is a good chance that it’ll survive in the future without any major black swan events."
The reason for the concern?
This would be true if the prior "stress test" periods started with situations similar to today's. But they didn't. So the assumption is dangerously invalid. Ignore the simulator.
An example: In 1929, usually the worst stress test, bond yields were over 5% and there was mild deflation, not inflation. (CPI fell steadily 1925-1933, total -30%). So real yields were closer to inflation + 6-7%.
And even at the 1929 bubble top, stocks were on many metrics very much cheaper than they are today.
e.g., the median P/S among non-financial firms in the S&P 500 and its predecessors never exceeded 2.0 before 2013 in all US history.
(looking at the 400 largest non-financial firms by market cap)
The tech bubble absolute peak of that median was 1.639 and the credit bubble absolute peak was 1.844
The latest figure is about 3.21
Now, with a bit of work, I think that sort of calculator can give some output that is of potential interest.
First, estimate the valuation level of bonds relative to their deep history, say 60+ years. Real bond yields versus their long run average.
Then, estimate the valuation level of stocks relative to their deep history, say 60+ years. Say, average of the results of a CAEY calculation and a P/S calculation.
Apply those haircuts to the starting portfolio size, then see the result. If the numbers are different, calculate a weighted average based on the likely stock/bond mix you'll be testing.
e.g., imagine you look at those numbers and estimate that both stocks and bonds are (say) 40% more expensive today compared to deep history.
So, the forecasts for a starting portfolio size of $60k entered in the simulator might give you a reasonable idea of the expected range of outcomes of a portfolio of mark-to-market value $100k starting today.
This is more or less consistent with this assumption: the market more or less immediately drops down to a valuation level that is historically typical of the past, then follows the range of outcomes seen in that same past. The reason is obvious: the historical distribution of outcomes of equities (or anything!) can never be thought of as representative of the future unless the starting situation is at least reasonable typical of the past.
This approach will still be too optimistic, though, because all historical equity return observations were in periods of gradually rising valuations. In a typical decade in teh past, around 1% of the observed return was from things simply getting more expensive.
If that trend does not repeat--and there is no reason to think it will--then all the historical returns are on average overestimates of the future likely returns even from the same starting valuation level.
Jim
