Implementing and Refining the “Spend Safely in Retirement Strategy”

A couple of years ago, we discussed a paper by Steve Vernon, Joe Tomlinson, and Wade Pfau, which looked at an assortment of retirement spending strategies and evaluated them based on several different criteria. The authors then put forth a strategy that they referred to as the “Spend Safely in Retirement Strategy,” which generally does a good job of satisfying the various (often competing) criteria.

Broadly speaking, the strategy involves creating two sources of retirement income:

1. A safe floor of guaranteed lifetime income. The authors refer to these as “retirement paychecks.” This includes Social Security, pensions, and annuity income. These retirement paychecks would be used to cover the necessities like housing, utilities, food, transportation, medical care, etc.
2. A liquid mutual fund portfolio, from which you pay yourself a “retirement bonus” — used for discretionary expenses. The level of spending from this portfolio varies with investment performance.

And again speaking broadly, the strategy has two steps:

1. Implement the safe floor of income. Usually this means delaying Social Security if you are single or the higher earner in a married couple. Sometimes it means delaying for the lower earner in a married couple as well. And sometimes it means buying an annuity for additional guaranteed income.
2. For the remainder of the portfolio (the “retirement bonus” portion), invest in a low-cost stock index fund or all-in-one fund (e.g., target-date fund, balanced fund, or LifeStrategy fund). Then use the IRS’s RMD tables to determine how much to spend from this part of the portfolio each year.

This strategy tends to work well as a rough-draft plan, for a few reasons:

• Satisfying basic needs via guaranteed income minimizes your exposure to investment risk, longevity risk, investment mistakes, cognitive decline, fraud, or mistakes that might otherwise be made after the death of the more financially knowledgeable spouse.
• To the extent that the guaranteed income is made up of Social Security, your exposure to inflation risk is minimized as well.
• Using the RMD tables for discretionary spending accounts for the facts that it is wise to adjust spending based on investment performance, as well as the fact that you can safely spend a greater percentage of the portfolio per year the older you are.
• The plan is reasonably simple and can in many cases be implemented without needing a financial advisor.

Real-World Implementation

But the basic, two-step plan described above (and in the original report) leaves an assortment of open questions. And when it comes time to actually implement the strategy in a real-world situation, you must come up with answers to those questions.

So I was happy to learn recently that the authors released a follow-up paper that addresses those real-life implementation questions one-by-one. (To be clear, the follow-up paper was published last year. I only recently learned about it though.)

The newer paper addresses questions such as:

• How would you implement the RMD portion of income before the normal RMD age? (In brief: use the same life-expectancy-based calculation that the IRS uses. The authors provide a table with per-year spending percentages.)
• How would you select an asset allocation for the RMD portion of the portfolio? (In brief: if your basic needs are completely satisfied by guaranteed sources of income, you can afford a stock-heavy allocation with remaining assets. Whether you want to use such an allocation is up to you and your preferences.)
• How can you plan for the fact that the portfolio-funded level of spending has to vary as the level of income from other sources (e.g., work income or Social Security) changes over time? (In brief: create a “retirement transition fund” — a portion of the portfolio that has been carved off and invested in something like a bond ladder that will be used to fund the additional spending over the years in question.)
• How can you plan for an uneven desired amount of total spending, such as a desire to front-load spending in the early years of retirement? (The authors propose a few options here. One such proposed method is to multiply the RMD for each year by a factor such as 1.25 or 1.5, which would increase spending early — and thereby result in less spending later, since you’d be spending a percentage of a portfolio that is smaller than it otherwise would have been. They run through a few examples of how such adjustments would have played out, given various assumptions.)

If you have the time, I’d encourage you to give the newer follow-up paper a read — or at least bookmark it for future reading. As I’ve written previously, I think the strategy that the authors describe is a great rough-draft approach to funding retirement spending (i.e., a sort of “cookie cutter” plan, which you can then adjust based on your own circumstances and preferences).