The withdrawal method or rate is just a tool to get a job done. As an engineer, physicist, and musician, I know that using the right tool for the job is vital and that working out what the 'job' actually consists of is an important first step.Simple question that I am sure will have many complex answers that will hopefully give people some good insight into how they want to manage their withdrawals.
Why is determining a withdrawal rate or method so difficult?
In the context of retirement, at it's most basic the 'job' of a portfolio is to provide income, so the first questions ought to be 'how much income is required in total and how much variability in that income can be tolerated?'
The first of those questions is then about planned or expected expenditure (itself not a trivial question to answer over a projected 30 or more years), while the tolerance to income volatility depends on the proportion of expenditure that can be considered 'core' (or 'essential') and the proportion that can be considered adaptive (or discretionary). While what spending is placed under which category will vary between retirees, the bare minimum for 'core' might consist of housing, heating, food, health, and transport. The 'core' spending (and, income required) would be intolerant of volatility, while there might be a great deal of tolerance in the 'adaptive' spending.
In matching volatility of income sources to the acceptable level of volatility in expenditure it can then be noted that:
a) social security is non-volatile in real terms
b) a DB pension may or may not be volatile in real terms (depending on inflation adjustments)
c) A collapsing TIPS ladder provides a non-volatile real income, but is limited in the time over which it can supply that income.
d) Nominal annuity income is volatile in real terms (and generally downwards)
e) Income volatility from a portfolio depends on the withdrawal method and (for some methods) the stock/bond allocation. Constant inflation adjusted withdrawals are non-volatile right up to the moment the portfolio is exhausted, percentage of portfolio methods (constant percentage, RMD, ABW, etc.) are as volatile as the portfolio value, while hybrid methods (e.g., Guyton Klinger, Carlson's endowment, Vanguard Dynamic, Bengen's floor and ceiling, portfolio smoothing, etc.) reduce the volatility but increase the chance of portfolio exhaustion compared to percentage of portfolio methods.
Taking an average sort of example. Average retired household expenditure in the US is around $60k, while average social security is around $34k for a couple. Let's assume they have a combined portfolio of $650k. Further assume their essential spending is equal to £34k and adaptive expenditure is $26k.
If they are happy for all of their adaptive expenditure to be volatile then they can pretty choose whatever withdrawal method they like.
If they are happy for only about half of their adaptive expenditure to be volatile, in essence this is moving half of the adaptive expenditure to core, then they could
a) build additional income flooring of $13k using a TIPS ladder (at a a cost of $300k, leaving $350k in their portfolio to supply the adaptive income and all income beyond social security after the ladder expires). The withdrawal method adopted then becomes immaterial (except after the TIPS ladder is spent down).
b) Use a hybrid method such as an income floor (of $13k) combined with a percentage of portfolio approach (e.g., see https://www.cfiresim.com/40da3d3f-9c94- ... 448161c792) or other types. The portfolio may become exhausted if future conditions are about 2.5 percentage points worse than historical markets (i.e., modelling higher inflation/lower returns by setting fees=2.5% in cfiresim as a stress test).
c) Combine aspects of a and b
a and c are essentially a floor and upside approach which has been discussed on these boards in the past (as a search will show).
I think the key thing, is that the decision of what withdrawal method to use follows from the role portfolio income is expected to play in supplying the overall income. The approach adopted will then be different for different retirees.
Sorry for a rather long answer to a short question!
cheers
StillGoing
Statistics: Posted by StillGoing — Fri Nov 08, 2024 4:33 am