Chapter 23: Turning a pile into a paycheck
For forty years the question was simple: how do I make this number bigger? Retirement flips it. Now the question is: how do I turn this pile into a paycheck that survives bad markets, taxes, health costs, and the genuinely good problem of living to 95? Growing money is mostly one decision repeated. Spending it is a coordination problem: withdrawals, taxes, benefits, and luck all interact. This chapter builds the paycheck.
Start with the income map
Before any withdrawal math, lay out two lists side by side.
What goes out. Split spending into essential (housing, food, insurance, health care, utilities: the bills that arrive whether markets cooperate or not) and discretionary (travel, gifts, hobbies: the spending you could flex in a bad year). The split matters because they get funded differently.
What comes in. In rough order of reliability: a pension if you have one; Social Security (Chapter 24 covers when to claim); cash reserves; your taxable brokerage account; traditional 401(k)s and IRAs (taxed as income on the way out); Roth accounts (tax-free); maybe an annuity; maybe home equity. Each source has its own tax treatment and its own best moment.
Carlos and Elena spend about $6,500 a month: $5,200 essential, $1,300 discretionary, or $78,000 a year. Coming in: Carlos's pension pays $2,300 a month ($27,600 a year), and they're delaying his Social Security to 70 (Chapter 24 explains why). Their accounts: $100,000 in cash, $300,000 taxable, $650,000 traditional, $150,000 Roth, about $1.2 million in all. The gap to fill until Social Security starts: $78,000 − $27,600 = $50,400 a year. The cash bucket alone covers two gap years; cash plus taxable covers the whole bridge. Once Carlos claims at 70 and Elena's benefit is flowing, pension plus Social Security will cover essentially all of their spending, and the portfolio's job shrinks from "paycheck" to "backup and upgrades."
That shape (heavy withdrawals for a few bridge years, light ones after) is common, and it's why a single flat withdrawal rate is too crude a tool on its own.
The 4% rule: a research result, not a promise
In 1994, William Bengen asked a blunt question of US market history: what's the most a retiree could have withdrawn in year one, then raised with inflation every year, without running out over 30 years, even retiring at the worst moments (1929, 1966)? The answer for stock-heavy balanced portfolios was about 4%. That's the 4% rule: $1 million supports roughly $40,000 in year one, inflation-adjusted after.
Use it as a sizing tool (it's why "25 times annual spending" defined FI in Chapter 22), but stress-test it before you bet a retirement on it:
- Longer horizons. Bengen tested 30 years. Retire at 55 and you may need 40.
- Lower future returns. History is a sample, not a ceiling or a floor.
- Fees. A 1% advisory fee turns a 4% withdrawal into a 5% drain on the portfolio.
- Care costs. Most people will need some long-term care assistance, and Medicare doesn't cover the custodial kind, a late-life spending spike the flat rule ignores.
- Early losses. The biggest threat of all, and it gets its own section below.
Sequence risk: when bad years arrive matters more than whether they do
This is the least intuitive fact in retirement finance. Two retirees can earn the same average return over 30 years and end in wildly different places, purely because of the sequence-of-returns risk: the order in which the good and bad years arrive once you're withdrawing.
Consider two retirees who start with $500,000 and withdraw $30,000 a year. Over 30 years, both experience identical returns (twenty-seven years at +9% and one ugly stretch of −15%, −4%, −10%), averaging 7.1% a year for both. The only difference: retiree A hits the ugly stretch in years 1–3; retiree B hits it in years 28–30.
Out of money in year 22. Early losses plus withdrawals ate the base that the later +9% years needed.
Same returns, same 7.1% average, same $30,000 withdrawals, and she ends with more than she started with.
While you're saving, order doesn't matter; the math commutes. Once you're withdrawing, every dollar sold in a down year is a dollar that can't recover. Early losses are permanent in a way late losses aren't.
Four defenses that actually work
1. A reserve of years, not just dollars. Hold 1–3 years of planned withdrawals in cash and short bonds (Chapter 3 vehicles). In a crash, spend the reserve instead of selling stocks at the bottom; refill it in recovery years. Carlos and Elena's $100,000 cash bucket is exactly this.
2. Flexible spending. The essential/discretionary split becomes a dial: a household that can drop spending 15% in a bad year has dramatically better survival odds than one locked at a fixed draw. Flexibility is the cheapest insurance in retirement.
3. Guardrails. Pre-commit the dial to a rule: skip the inflation raise (or cut 10%) after a bad year, allow a modest raise after a strong run. Written rules beat in-the-moment judgment. That was Chapter 14, and it's truer at 70 than at 40.
4. Floor and upside. Cover essential spending with guaranteed income (pension, Social Security, possibly an annuity) so markets can only ever interrupt the fun, never the groceries. Then invest the rest for growth with a clear conscience.
Build the floor first: lock enough guaranteed income (pension, delayed Social Security, maybe a simple annuity) to cover essentials for both lifetimes. Let the market fund only what you could survive flexing.
Annuities, honestly
An annuity is a contract with an insurer: you hand over a lump sum, they pay you income. The simplest form, a single-premium immediate annuity (SPIA), does one real thing well: longevity pooling. People who die early subsidize people who live long, so the insurer can pay every survivor more per year than they could safely withdraw alone. If you fear outliving money more than you desire a bequest, that's genuine value.
The honest cautions, per the SEC's Investor.gov: many annuities sold are not simple SPIAs but complex variable or indexed products with layered fees, surrender charges (penalties for taking money back out, sometimes for 7–10 years), and caps that quietly trade away upside. The guarantee is also only as strong as the insurer behind it, backstopped by state guaranty associations with limits. So: simple products, top-rated insurers, and only enough to finish the essential floor.
And before buying any annuity, price the one the government already sells you: delaying Social Security from 62 to 70 raises an inflation-protected, survivor-protected payment by roughly 77% (70% of the full benefit versus 124%; Chapter 24 does the math). For most retirees that's the best annuity money can buy, and it requires no salesperson.
Where people go wrong
- Retiring on an average. "Markets return 7%, I'll withdraw 5%" ignores sequence risk entirely. Averages don't pay bills; specific years do.
- Keeping the accumulation portfolio. A 90% stock mix that was perfect at 45 has no shock absorber for the bridge years. The reserve of years has to exist before the last paycheck, not get built during the first crash.
- Treating the plan as set-and-forget. Withdrawal plans are recalculated, not declared. One honest checkup a year (spending, balances, the guardrail rule) is the whole maintenance schedule.
- Buying complexity to avoid a simple fear. An eight-rider indexed annuity is rarely the answer to "I'm scared of running out." A bigger cash buffer, a flexible budget, and a delayed Social Security check usually are.
Key takeaways
- Retirement flips the problem from growth to coordination: map essential vs. discretionary spending against every income source before touching withdrawal math.
- The 4% rule is a useful historical sizing result, not a guarantee; stress-test it for longer horizons, lower returns, fees, and care costs.
- Sequence risk is real: identical average returns in a different order took one $500,000 retiree to zero in year 22 and another to $1.22 million in year 30.
- Defend with a cash reserve of years, flexible spending, written guardrails, and a guaranteed-income floor under essentials.
- Annuities pool longevity usefully but come wrapped in complexity and fees, and delaying Social Security is usually the cheapest annuity available.
Sources: Investor.gov: Asset allocation · SSA: Delayed retirement credits · ACL: How much care will you need? · Bengen (1994), "Determining Withdrawal Rates Using Historical Data"