Retirement Income Planning

Retirement income planning is the work of converting a lifetime of savings, Social Security, and any pension into spendable, sustainable income — and of deciding in what order to make the decisions that produce it. It is the umbrella over a set of narrower questions that each have their own answer: how much can be withdrawn each year (the safe withdrawal rate), which accounts to draw it from (the withdrawal strategy), how the portfolio is segmented so a withdrawal in a downturn does not force a stock sale (the bucket strategy), when to claim Social Security, whether to buy guaranteed income, and how required minimum distributions eventually take some of the choice away. What belongs to this page rather than to those is the sequence, the risk framework, and the transition problems nobody else owns.

The order of operations is the content. Most retirement income mistakes are not wrong answers, they are answers produced in the wrong order.

1. Inventory the guaranteed income you already have. Social Security for each spouse at various claiming ages, any pension and its survivor option, any annuity already owned. This comes first because it determines the size of every problem that follows.
2. Quantify the spending the portfolio must actually cover. Not a replacement percentage of your old salary — a real estimate, split between essential and discretionary, in today's dollars, including the costs that rise in retirement (health care, travel early on) and the ones that fall (payroll tax, commuting, saving itself).
3. Choose a sustainable withdrawal rate for the gap between the two. That is the safe withdrawal rate question, and it has its own page.
4. Decide which accounts to draw from, and in what order. This is a tax problem, not an investment problem, and it is the withdrawal strategy question.
5. Stress-test and revisit. Test the plan against a long life, a bad first decade, higher inflation, and a survivor scenario — then check it annually, because a plan that is never revisited is a forecast, not a plan.

The four risks. Retirement income planning exists because four risks all land on the same portfolio. Longevity risk — the money must last an unknown number of years. Sequence-of-returns risk — a bad first decade of returns does permanent damage that a good average cannot undo. Inflation — a fixed income stream loses purchasing power every year, and over thirty years that compounding is severe. Health and long-term care — the largest and least predictable late-life expense, addressed partly through long-term care insurance and partly through reserves. Each has its own page; what matters here is that a plan addressing only one of the four is not a plan.

Floor versus probability: two defensible frameworks. The floor approach covers essential spending with guaranteed income — Social Security, a pension, and if necessary an annuity — and treats the remaining portfolio as discretionary, to be spent flexibly. Its appeal is that the worst case is survivable by construction, and it tends to make retirees more willing to spend. The probability approach keeps everything in one portfolio and asks what withdrawal plan succeeds in a high percentage of simulated futures, using Monte Carlo simulation. Its appeal is flexibility and the likelihood of leaving more behind. Neither is objectively correct: the first buys certainty with liquidity and upside, the second keeps liquidity and upside at the cost of living with a probability. Which suits a household is a temperament question as much as a financial one, and hybrids are common.

The transition problems. Three specific issues appear only at the moment of retiring, and none of them is a portfolio question. First, the health insurance bridge — anyone retiring before 65 needs coverage until Medicare begins, and its cost is often the single largest surprise in an early-retirement plan, with the added complication that the premium subsidy available on the individual market depends on the income the retiree can partly control. Second, the low-tax window between the last paycheck and the start of Social Security and required minimum distributions, which is usually the most valuable tax-planning opportunity of a lifetime and closes permanently. Third, the shift from accumulation to decumulation: for thirty years the job was to save more and stay invested, and suddenly the job is to spend down an irreplaceable asset on an unknown schedule. That is a genuinely different discipline, and it is why savers who did everything right can still be poor decumulators.

One structural note about who builds the plan. Retirement income is the point in a financial life where product sales concentrate, because it is the moment a large balance becomes movable and the buyer is most anxious about running out. Annuities, insurance-based income strategies, and managed portfolios all get pitched hardest here, and several of them may be genuinely appropriate. That is precisely the argument for having the plan built by someone whose compensation does not change depending on which of them you choose.

A hypothetical example of steps 1 and 2, which is where the process either gets a grip on the problem or does not.

Ruth and Tom are both 64 and want $96,000 a year before tax. Their guaranteed income, if they wait until 67, would be a combined $52,000 from Social Security plus a $9,000 pension of Tom's — $61,000 of income that arrives whether markets cooperate or not. So the portfolio's real job is $35,000 a year, not $96,000, which is a far smaller and more solvable problem than the one they thought they had.

Then the piece almost every do-it-yourself plan misses. They are retiring at 64 and their Social Security does not start until 67, so for three years the portfolio must cover the whole $96,000 — about $288,000 of front-loaded withdrawals before the guaranteed income ever appears. That bridge is not an inconvenience, it is a structural feature of their plan: it shapes how much cash they need on hand, it is exactly when sequence-of-returns risk hits hardest, and it is also the low-tax window in which Roth conversions are cheapest. Three separate decisions all live in those three years.

Only after this is quantified do the narrower questions have answers worth computing: what withdrawal rate the remaining portfolio supports, which accounts to draw from, how to segment the portfolio, and whether any of the $35,000 gap should be covered by guaranteed income rather than by the portfolio. (All figures hypothetical and illustrative.)

Where each decision is covered

• Safe withdrawal rate — how much per year the portfolio can support, and the research behind the answer.
• Withdrawal strategy — which accounts to draw from in what order, and the tax consequences of getting it wrong.
• Bucket strategy — how to organize the portfolio by time horizon so a downturn does not force a stock sale.
• Social Security retirement benefits — claiming ages and how the decision changes lifetime income.
• Annuity, and specifically the immediate annuity and the qualifying longevity annuity contract — how to buy guaranteed income, and what it costs.
• Required minimum distribution — when the tax code starts forcing withdrawals and removing choice.
• Sequence of returns risk and longevity risk — the two portfolio risks the whole structure is built around.
• Monte Carlo simulation — how a probability-of-success plan is tested.
• Long-term care insurance — the late-life expense that most often breaks an otherwise sound plan.

Pros (of treating it as an explicit process)

• Sequencing the decisions prevents the most expensive category of error: solving a narrow question well before knowing what problem the portfolio actually has to solve.
• Quantifying guaranteed income first usually shrinks the apparent problem dramatically and makes the rest tractable.
• It surfaces the time-limited opportunities — the low-tax window before Social Security and required minimum distributions — while they are still open.
• Naming all four risks stops a plan from being over-engineered against one and silently exposed to the others.

Cons (and the limits of planning)

• Every output depends on inputs nobody knows: lifespan, returns, inflation, tax law, and future health costs.
• A high probability of success is a statistical statement, not a promise, and it can create false confidence.
• The process demands a genuine spending estimate, which most households have never produced and find uncomfortable to build.
• Plans go stale. Without an annual revisit, a carefully built plan becomes a document about a version of your life that no longer exists.
• This is the part of financial life where sales pressure is heaviest, so the quality of the advice matters more here than almost anywhere else.