I have a very granular budget entered into Pralana Online, with one category as a possible discretionary vs non-discretionary expense. So I believe my expenses are well-represented, including some future cars, etc. I'm not worried about long term care at this time, because my plan is to use my house for that, although I may start a new scenario in the near future for preferred in home care. My purpose right now is to model the effect of possible Social Security income decreases in 2033. My assumption is a reduction of 23%. That irritates me, because it's more than half of what I am defining as discretionary. I'm nearly 68, with a plan running to age 100.
So using a spending strategy of "Specified Expenses Only", I have 2 scenarios, one with my expected Social Security, and one with the reduction. Then I varied whether my "discretionary expenses" were essential, or not, in the set up. I ran a Monte Carlo and an Historical Analysis on each scenario. In both cases Historical Analysis, starting in 1928, did better, coming in at 100% for both scenarios, both expense situations. Monte Carlo was a little worse, coming in at 95/96% if SS stayed the same, and 89/90% if SS is reduced. So I may be ok with still spending my discretionary $$, and I may rework my budget even more to decide what is truly discretionary vs sort of variable. That's a topic for another day.
I still can't get through my head why Historical analysis always seems to come out better for me than Monte Carlo. I think I've asked this before, but can't look right now.
Regardless, my main question is, what spending strategy would generate a more conservative approach than Specified Expenses? I'm not concerned with leaving a legacy to my sons, but more concerned that I don't cause problems for them financially at the end of my life.
Really appreciating this very powerful tool.
Some version of an actuarial method might be considered more conservative, as it allows variable withdrawals each year beyond expenses to adjust up and down with how the market does. Pralana contains one version of this method ("actuarial"), but VPW at bogleheads is another popular and even simpler method. The root idea is that each year you take your remaining portfolio (after future ss added and future expenses deducted) and divide it roughly into the number of years left in the plan. That's a bit of a simpliciation, but captures the basic idea of "lifecycle withdrawals". What I like about it is that if the market does well you take out more and if does worse you take out less, with risk spread over all remaining years. So rather than a "fixed" constant amount that gets you in a bind or leaves you too much, you'll find that amount goes up and down, but ends up leaving you with zero. That is a plus and a minus depending on your ability to handle variable withdrawals.
I still can't get through my head why Historical analysis always seems to come out better for me than Monte Carlo. I think I've asked this before, but can't look right now.
One possible reason might be that your future estimates of returns on equities might very well be lower than the historial average. Many people, myself included, predict only 5% or so real returns for stocks compared to a 6.7% average historically.
The root idea is that each year you take your remaining portfolio (after future ss added and future expenses deducted) and divide it roughly into the number of years left in the plan.
That tactic gives me a panic attack as the thing I find folks enter most inaccurately is their date of expiration 🙂 Much better if you're accounting for longevity risk and saying some thing like age 100, but far too many come to us as clients and say, "Put down 70, lots of people in our family die at 70..." We don't want anyone working as a Walmart greeter at age 80 unless it's truly their passion 🙂
The root idea is that each year you take your remaining portfolio (after future ss added and future expenses deducted) and divide it roughly into the number of years left in the plan.
That tactic gives me a panic attack as the thing I find folks enter most inaccurately is their date of expiration 🙂 Much better if you're accounting for longevity risk and saying some thing like age 100, but far too many come to us as clients and say, "Put down 70, lots of people in our family die at 70..." We don't want anyone working as a Walmart greeter at age 80 unless it's truly their passion 🙂
1) I don't disagree! I have my age 102 (wife age 95) as the assumed last year of our plan. I use the 25th probability that one of us will still be alive from actuarial tables. However, this needs to be re-evaluated each year, and when we get to 80 or so an annuity might be considered if it's iffy.
2) The actuarial method is better suited to varying longevity, because it continually adjusts. Think of the RMD table: each year you don't die it realizes you have a higher chance of living more years than you did last year. And essentials are covered by safe assets, so "worst case" the discretionary non-essential assets run down to low amounts because you've underestimated your longevity early on or the market did poorly.
