Non-investing personal finance issues including insurance, credit, real estate, taxes, employment and legal issues such as trusts and wills
In addition to the inflation risk and spending many extra years working relative to a traditional plan that is the same or less risky, I think using current life expectancy numbers for decades in the future is far more speculative than most people assume. Life expectancy for an 80-year-old could be vastly higher in 20-30 years than it is today.
Agreed. At least scenarios, stochastic is an improvement. I'd take it further and realize that historical data is not applicable as a starting point. It might be ok to drift toward historical data later on in the projection (maybe around year 10 to 15).dbr wrote: ↑Wed Oct 13, 2021 8:36 amSpreadsheet calculations using assumptions about inflation and return can be very wrong if those assumptions don't match the reality of future events. It would be much better to examine probabilistic ranges of inputs and see how possible success or failure might be.reln wrote: ↑Wed Oct 13, 2021 7:49 amIn this hypothetical example in which a retiree has sufficient cash to support a 40 year personal-inflation adjusted spending, then why not just hold cash? By the premise it would work out for the person. Back to reality, I think risk exists by both holding financial assets and by not holding them. You may not worry about stocks if you don't hold them but you may fall into worrying for not having them at some point. Also, you can learn to hold them and not worry about them.Gnomon wrote: ↑Sat Oct 09, 2021 5:41 pm In theory, if a person has enough saved at retirement to cover their inflation-adjusted expenses over their estimated lifetime, they would not have to take any stock market risk with their savings, and would have no sequence of returns risk since there is no reliance on market returns.
That zero market risk approach always seemed appealing to me and that's what I have been aiming for. I'd like to be done with worrying about markets, especially in the early retirement years.
In the most basic form (i.e. assume zero social security benefit and zero return on your cash portfolio), if you retire having $40k in annual living expenses, a 40 year retirement timeline, and assume 2.5% inflation, then I think you can use the FV() function in excel to figure out what your starting portfolio size needs to be:
in Excel: =FV(2.5%, 40, -40000)
And that equals $2,696,102, or 67.4x initial expenses. That's not an attainable number for most people. But…if you include beginning a social security benefit at age 67 and say 1.25% return on cash via e.g. 5 year CDs (which is lower than past averages), the final number becomes more attainable (as long as you had a full SS work history and a SS benefit that initially covers a good portion of your living expenses).
Example: if you retire at age 60 and your SS benefit at age 67 is $38,725 annual (~$3225/mo) and is COLA-adjusted 1.4%, and you earn 1.25% APY on your cash portfolio for the duration of retirement, and the starting value of your portfolio is $1M with $40k annual expenses in the 1st year (i.e. 25x expenses), you end up with $197.5k remaining at the end of a 40 year retirement at age 100, without any stock/bond market exposure.
Or, $393.6K left if you die at age 95, or $539.6k left if you die at age 90. And of course even more headroom if your initial cash portfolio is larger.
It even works pretty well if you begin taking your SS benefit at age 62 at $27,276 annual ($2273/mo), otherwise same parameters as above, although you would run out of money at age 98.
Or with a $1.5M portfolio and beginning SS at 67, else same as above, your initial annual expenses could increase to $52k (30% higher than the baseline case) and you’d run out of money right at age 100.
Is there anything fundamentally wrong with this approach of taking zero market risk during retirement, if the numbers work out as described in the scenarios above? (Granted, it’s possible I’ve made a mistake in my spreadsheet calculations.) Not having to worry about market returns, crashes, or recovery time for the duration of a 40 year retirement does have appeal. I’d be interested to hear views on this, or additional factors to consider if taking this approach.
I would have more confidence in even a historical model like FireCalc (and there are many others) than in constructing spreadsheets with dubiuous and non-varying assumptions. Just for a starter if one does run FireCalc with cash (1 month Treasuries) for 40 years, the starting portfolio needs to be 67x expenses to get no failures out of 40 historical trials. With "cash" being five year notes you need 50x expenses out of 111 trials to get zero failures. Note Firecalc does have some limits regarding what is cash and what historical period is being represented. Other tools may give a better view.
The point is that it (all cash) can be done, but the situation is extreme when it does not need to be. Add 20% stocks to that 50x example and you can retire at about 35x expenses for 40 years. If you look at a typical efficient frontier for stocks and bonds 20/80 gives more return at actually less risk than 0/100.
It is hard to understand what the motive would be for supporting a long retirement with no stocks as the uncertainties are actually greater and the risk larger.