This post will provide some background on the benefits of establishing separate reserves dedicated to funding different categories of expenses in the future, and why we believe the Actuarial Approach generally accomplishes this task much better than the approach suggested by Mr. Kitces in his post.
Background
For several years we have encouraged our readers to separate the total present value of their expected future expenses into various sub categories, such as:
- Present value of recurring essential expenses
- Present value of non-recurring essential expenses,
- Present value of recurring non-essential expenses, and
- Present value of non-recurring non-essential expenses
The primary reason we encourage you to separate non-recurring expenses from recurring expenses is because non-recurring expenses are generally not incurred every year, and it would be unnecessarily expensive to reserve for them assuming they were. For example, if your home mortgage will be paid off in five years, you should plan to incur this non-recurring expense for the next five years only, not your entire lifetime planning period.
As discussed most recently in our post of January 16, 2019, we also suggest that you consider categorizing your recurring and non-recurring expenses as either “essential” or “non-essential” to provide you with information regarding the relative size of your asset reserves dedicated to funding these types of expenses. This information can also be useful when developing an investment strategy and in making spending decisions. In our post of April 10, 2019, we discussed why you might want to create separate “floor” and “upside” portfolios dedicated to funding essential and non-essential expenses, with very different investment strategies applied to the two types of portfolios.
We have always left it up to our readers just how to categorize their future expected expenses. We have indicated that it may be reasonable to assume different rates of future increases for different types of expenses (like essential health care expenses for example). Our Actuarial Budget Calculator (ABC) for single retiree workbook has a separate “Budget by Expense” tab for the purpose of separating the present value of future expected recurring expenses into three types of expenses, but there are no reasons why additional categories could not be added.
Mr. Kitces’ Proposed Approach
In his post, Mr. Kitces argues:
“In essence, separating “essential” vs. “discretionary” expenses by looking at categories of spending doesn’t necessarily work. Instead, the better approach is perhaps to segment spending within each category into the “Core” expenses that form the nucleus of the household’s lifestyle, from the truly discretionary (and more easily adaptable) expenses within each category (from the upscale restaurants to the designer clothes). In other words, there’s a potential layer of Core vs. Adaptive expenses in every spending category.”
Mr. Kitces argues that this distinction is not just semantics, and we are not inclined to argue with him on this point. In our opinion, however, the more important take-away from Mr. Kitces’ post is not the process used to select the specific expenses to segment, but rather how the appropriate reserve/bucket to fund that particular category of expected future expenses is established once the categories of expenses have been selected in a manner consistent with retiree spending goals and tolerance for risk.
Mr. Kitces develops his reserve for core expenses by subtracting expected Social Security benefits from expected core expenses and dividing the result by 4% (based on the 4% Rule). The annual core expense budget, then, would be equal to Social Security benefits plus 4% of such fund at time of establishment and would be increased with inflation each subsequent year.
His reserve for adaptive expenses is established by dividing expected adaptive expenses by 7%, which according to Mr. Kitces is the 4% Rule adjusted to be consistent with a 20-year period as opposed to the normal 30-year period underlying the 4% Rule. It appears that the annual adaptive expense budget would not be increased by inflation each year after the year of establishment of this fund, but would remain at 7% of the initial adaptive expense fund in each future year.
We personally have no problem if you prefer the terms “core” and “adaptive” rather than “essential” and “non-essential.” We do believe, however, that it would be beneficial for these two types of expenses to be further segmented into recurring and non-recurring core/adaptive expenses, and we believe using present values to develop the required reserves for these expense categories (and other expected expenses) is superior to using variations of the 4% Rule proposed by Mr. Kitces.
Using the Actuarial Approach
First, we are happy to report that when we used our Actuarial Budget Calculator (ABC) for Retired Couples, we were able to fairly closely reproduce Mr. Kitces’ numbers for the hypothetical couple discussed in his post. When we used the default assumptions (and 33% reduction in spending upon the first death within the couple), we determined that accumulated savings of $520,000 plus the Social Security amounts (including the 5 year deferred commencement) assumed for the hypothetical couple by Mr. Kitces, would produce an initial annual recurring spending budget of $54,433 (vs. Kitces’ $54,400). This compares with Mr. Kitces’ Core fund of $485,000 ($360,000 plus Social Security Bridge of $125,000).
If we assume a 4% discount rate, no future inflation increases and a 20-year payout period, we confirm that a $350,000 Adaptive fund will generate about $24,763 per annum, vs. the $24,800 amount determined using Mr. Kitces’ proposed approach.
By using his approach, Mr. Kitces’ hypothetical couple has decided to “front-load” their adaptive spending by treating all such spending as non-recurring expenses expected to last exactly 20 years (with no such adaptive expenses anticipated for the “no-go” years after age 85). As discussed in our post of August 12, 2018, treating expenses as non-recurring rather than recurring is one way to front-load spending. We do believe, however, that there may be situations where individual non-recurring or adaptive expenses may be expected to last either longer or shorter than 20 years, and should not simply be all lumped together as one expense.
And while the Actuarial Approach can fairly closely reproduce Mr. Kitces’ results for his relatively simple example, it can also develop reserve buckets and spending budgets for much more complicated situations. Unlike Mr. Kitces’ approach, it can handle:
- Couples budgeting for couples with different ages
- Income from other sources, such as part-time employment, fixed and indexed immediate life annuities, pension benefits and deferred life annuities with different start and stop dates.
- Expected and unexpected non-recurring expenses with different start and stop dates.
Conclusion
We agree with Mr. Kitces that “A fundamental step to the process of budgeting and making responsible spending decisions is to separate “needs” from “wants.” We also agree that “clarifying the relative amount of Core vs. Adaptive spending also makes it easier to determine how much in retirement assets should be allocated to each bucket, how the buckets themselves should be invested, and whether the retiree actually has “more than enough” and could allocate some additional reserves to support even higher spending.”
We appreciate the fact that Mr. Kitces has employed the 4% Rule in his approach and has not implied that better results could be achieved through the use of Monte Carlo modeling. We believe, however, that the Actuarial Approach and its actuarial balance sheet comparison of the present values of assets and the present values of spending liabilities is a more robust approach for both simple and more complicated planning situations than the approach advocated by Mr. Kitces.
