Turning Your Savings Into a Paycheck

For thirty years the job was to add. Then one morning it is to subtract, and almost nothing that made you good at the first phase helps with the second.

Accumulation forgives mistakes: a bad year gets absorbed by the contributions still arriving and the decades still ahead. Decumulation does not, because a withdrawal taken during a downturn is permanent — the shares sold to fund it never participate in the recovery. The other half of the difficulty has nothing to do with markets: disciplined savers are frequently terrible spenders, and a withdrawal plan mostly gives permission.

The decisions also interlock. Which account you draw from changes your taxable income, which changes your Medicare premium two years later, which changes what a Roth conversion costs, which changes what your heirs receive. This is where coordination stops being abstract.

Which account you draw from first is answered by comparing what a dollar costs to withdraw from each treatment — this year and across two lifetimes.

The conventional sequence — taxable, then tax-deferred, then Roth — is a defensible starting point and frequently wrong in the details, because the years between the end of employment income and the start of required distributions are often the cheapest tax years of a life. Drawing only from taxable accounts wastes them, and the bill arrives later — when required distributions, Social Security, and a surviving spouse’s filing status land in the same year.

Filling those years on purpose is the opportunity: partial Roth conversions, realizing long-term gains while the applicable rate is low, and spreading distributions rather than bunching them — paying tax at a rate you choose instead of a rate the calendar chooses for you. Each has to be modeled against the thresholds in the way: bracket boundaries, capital gains breakpoints, IRMAA, the net investment income surtax, and whatever your state does. Those figures change, which is why the answer is a projection each year rather than a rule. Tax drag explains why the difference compounds.

Take two retirees with identical portfolios, identical spending, and identical average returns over twenty-five years. The one who gets the poor years first runs out of money; the one who gets them last leaves an estate.

Spending is denominated in dollars, but withdrawals are executed in shares. When prices are down, funding the same dollar of spending requires selling more shares, and those shares are permanently removed — so the recovery arrives for a smaller base. The loss is not the decline itself but its interaction with the withdrawal, locked in at the moment of the sale.

The defenses are structural rather than predictive, because nobody knows whether the poor years come first. What you control is whether a poor year forces a sale: a reserve sized in years of spending, income that arrives without a transaction, and spending that can flex. The third does more work than the other two combined and costs nothing to arrange in advance.

There are two ways to generate a retirement paycheck: hold assets that pay you — rent from real property, interest from credit, distributions from an operating asset — or hold assets that appreciate and sell pieces of them. The first does not require accepting a price on a particular day — the vulnerability sequence risk exploits.

That structural advantage gets oversold. A distribution is not free money: it can be a return of your own capital, it can be cut when the underlying cash flow falters, and it is frequently taxed as ordinary income rather than at capital gains rates. Reaching for yield to avoid the discomfort of selling takes on risk you did not intend to hold.

Real assets and private credit can play a role in the income tiers of a plan — different payers, different drivers, cash flow that does not depend on a market being open. They also lock capital up, the opposite of what a spending plan wants. The liquidity tiering in portfolio construction reconciles those two facts, and what alternatives are covers the categories. I need more income comes at the same question from the other side.

Social Security is longevity insurance priced by statute. Delaying produces a larger inflation-adjusted benefit for life and a larger survivor benefit for whoever outlives you, paid for by spending your own assets during the wait. The question is which risk you would rather carry.

The inputs that move the answer are health and family history, the gap in ages and earnings between spouses, whether a survivor’s income needs protecting, and whether the withdrawals that fund the wait push you into a higher bracket or a Medicare surcharge. For a married couple the survivor benefit frequently dominates — the higher benefit is the one that continues, which makes the higher earner’s claiming date a decision about the widow or widower.

Break-even analysis optimizes for the average lifespan, and the case that hurts is living a long time with a smaller benefit. I would rather insure the outcome that is unrecoverable than optimize the one that is merely suboptimal. The earliest, full, and maximum claiming ages are set by statute and have been adjusted before, so a plan built years in advance needs re-checking.

1. 1

   Get the spending number honest

   Not a budget — the actual figure, from a year of bank and card activity, split into what must be covered every month and what can flex. Most people are wrong in the same direction, and every decision downstream inherits the error.

2. 2

   Cover the fixed layer with the most reliable sources

   Social Security, any pension, and the most dependable cash flow get matched against the spending that cannot be reduced. What remains is what the portfolio has to produce — usually less than the gross figure suggests.

3. 3

   Size the reserve in years, not dollars

   The reserve exists so that a bad market never forces a sale. Sized in years of the portfolio-funded portion of spending, it becomes a number you can evaluate rather than an amount that felt comfortable.

4. 4

   Choose the withdrawal sequence with your CPA before the year starts

   A projection built in the second half of the year, covering the coming year, turns sequencing into an executed decision. Ordering the same withdrawals in December, after the income is recorded, forfeits most of the benefit.

5. 5

   Revisit annually and after anything material

   Tax law moves, markets move, health changes, and a spouse’s death changes filing status and brackets in the same year it changes everything else. A withdrawal plan is a position you adjust, not a document you file.