This has some ideas similar to Hebeler's Auto Pilot Method, except he split withdrawals 50/50 between the 4% rule and the RMD Method.In the recent, rather long and now locked thread, viewtopic.php?t=429052, there was some interesting discussion about flexibility in spending and consequently in portfolio withdrawals. I thought it might be interesting to look at the effects of adopting flexibility in portfolio withdrawals since there are a range of outcomes, some obvious and others not so obvious.
1. Most obviously, flexible withdrawals result in variable portfolio income.
2. Greater flexibility means, for a given initial withdrawal, the portfolio is more robust against prematurely running out of money if that initial withdrawal was poorly chosen.
3. Alternatively, looking at Point 2 from a different perspective, for a given level of portfolio robustness, more flexibility allows for a higher initial withdrawal.
4. Greater flexibility tends to leave a greater legacy in poor retirements, but a smaller one in good retirements. In other words, flexibility helps preserve the portfolio when necessary and spends it down when possible.
5. Not every retiree will be willing, need, or be able to incorporate the same amount of income variability into their retirement plan. Furthermore, for a given retiree, their willingness, need, or ability to adopt income variability may change as their retirement progresses.
In the rest of this post and those following, I will try provide evidence from backtesting to support each of the above propositions using an example flexible (dynamic) withdrawal strategy. Hopefully, quantifying the effects of flexibility, will also provide a pointer to those wanting to use an ad-hoc approach as to the amounts they might need to cut from their spending in the bad times or, more optimistically, be able to add in the good times!
In all of the backtests, I will use a 60/40 TSM/TBM portfolio (real return data taken from the Simba database), rebalanced annually and annual withdrawals taken at the beginning of each year. The effect of fees in excess of those accounted for in the database and taxes have been ignored. In each case, I have run the simulation for a planned 40 year retirement.
In order to provide a well-defined amount of flexibility, the total portfolio withdrawal consists of two components. A portion, F using a percentage of portfolio strategy and a second portion, 1-F using constant inflation adjusted withdrawals. For example, let’s assume a $1m portfolio, a 4% initial withdrawal and F=30% (i.e., 0.3). The initial withdrawal of $40k would consist of $28k ($1m*0.04*0.7) that will subsequently be inflation adjusted and a percentage of portfolio component of $12k ($1m*0.04*0.3). If inflation was 3% and the portfolio fell to $800k, then the second total withdrawal would consist of an inflation adjusted component of $28k*1.03=$28.84k, while the percentage of portfolio component would be $800k*0.04*0.3=$7.2k for a total of $28.84k+7.2k=$36.04k.
If you are familiar with a large number of withdrawal strategies, you might notice that this is similar to Carlson’s endowment formula (e.g., https://www.retirementwatch.com/how-to- ... st-forever) with the important difference that the inflation adjusted withdrawal is based on the initial withdrawal rather than the previous year’s total withdrawal being used as the basis of the next inflation adjusted withdrawal. For the backtests, I have used an amortization approach for the percentage of portfolio component such that the first value is approximately 4%, i.e., pmt(2.68%,40,-100,0,1) rather than a fixed percentage of portfolio, in order to be more effective in spending the portfolio down in good retirements.
While this approach is relatively easy to understand and implement and the flexibility is proportional to F with no hysteresis, it has the disadvantage (as we will see) that, unlike Carlson’s approach, the portfolio can still be prematurely exhausted. I note that there are other dynamic strategies that allow the flexibility in portfolio income to be parametrically adjusted which I could have used instead (e.g., Carlson’s endowment formula, Bengen’s floor and ceiling, Vanguard Dynamic Spending Strategy, portfolio value smoothing, etc.).
In the next post, we will begin testing the propositions.
cheers
StillGoing
BTW, for those who like amortization, the article below that describes auto pilot refers to it as "the planner method" and gives it a brief discussion in his list of withdrawal methods.
https://www.marketwatch.com/story/put-r ... 2013-07-24
For the method described in the OP, why wouldn't the fixed portion be a TIPS ladder + withdrawals based on NPV of future SS streams or SS Bridge via a short TIPS ladder? This allows you to look at flexibility in spending that's due to stock only (or any other risk AA) while maintaining a fixed portion that isn't subject to running out of money prematurely like the 4% or any other SWR percentage would.
Cheers.
Statistics: Posted by dcabler — Sat Apr 13, 2024 4:37 am — Replies 2 — Views 158