How to Invest for Retirement Income: Four Approaches and Their Tradeoffs

There's no single best way to turn a portfolio into a paycheck. Each sound approach handles some problems better than others: steady income, a bad market early on, a long life, access to your money, growth, and whether you'll stick with the plan when markets get rough.

It helps to separate three decisions that often get lumped together:

1. How the portfolio is invested. The mix of stocks, bonds and cash.
2. How much you can withdraw each year. The rule that sets this year's spending.
3. Where the spending cash actually comes from. Which holdings you sell or draw down when the bills arrive.

Many arguments about retirement income come from mixing these up. Buckets mostly answer the third question, guardrails the second, and an annuity changes all three. (Which accounts to draw from for tax purposes is a fourth decision, covered in our article on tax efficiency in retirement.)

One couple, four approaches

To compare the approaches fairly, here's one hypothetical couple. They're both 67 and newly retired. They spend $120,000 a year, including taxes. About $90,000 of that is essential (housing, food, insurance, health care) and $30,000 is discretionary (travel, gifts, projects). Their combined Social Security is $65,000, which leaves a $55,000 gap for the portfolio to fill.

Their portfolio is $1.5 million, invested about half in stocks and half in bonds and cash. Their starting withdrawal rate is $55,000 ÷ $1,500,000, or about 3.7%.

For context, Morningstar's 2025 research put the highest "safe" starting rate at 3.9% for someone who wants the same inflation-adjusted amount for 30 years, with a 90% chance of not running out, in a portfolio of 30% to 50% stocks. That estimate rests on Morningstar's own assumptions and isn't a promise, but this couple isn't starting from an aggressive position.

To test each approach, assume a bad first year: after markets fall and they take their $55,000, the portfolio is down to $1.2 million. That's hypothetical, not a forecast. The annuity approach starts from a different balance, so it's handled separately below.

Why the first years matter so much

A loss early in retirement does more damage than the same loss later, because the money you withdraw while prices are down isn't there for the recovery. That's sequence-of-returns risk, and Vanguard notes it can permanently impair a portfolio. Morningstar found that an all-stock portfolio supported only a 3.4% safe starting rate, partly for this reason. Every approach below is, in part, an answer to this problem.

1. A total-return portfolio, rebalanced

The couple holds a diversified 50/50 mix and funds spending from interest, dividends and sales as needed, rebalancing back to the target each year. Assume that in the bad year stocks fell more than bonds, leaving stocks below half the portfolio. Rebalancing then points spending toward bonds, and some bonds may be sold to buy stocks while they're cheap. In a year when stocks did better, the same rule would trim stocks instead, and interest and dividends may cover part of the withdrawal, so selling bonds isn't always required. It provides a disciplined framework for making decisions without reacting to every market movement.

Two points are often misunderstood:

• Rebalancing is not a withdrawal rule. It decides where this year's money comes from, not how much you can safely take. The couple still has to decide whether next year's $56,650 (the $55,000 plus 3% inflation) makes sense from $1.2 million. That's now 4.7%, not 3.7%.
• Dividends and interest aren't a safer paycheck. They're part of the portfolio's total return, just like price gains. Vanguard's research on total-return investing warns that tilting toward high-yield stocks or bonds to "live off the income" can quietly add concentration and credit risk. What matters is total return and the withdrawal amount, not what form the money arrives in.

2. Guardrails: spending that adjusts

Guardrails, or "dynamic spending," set rules in advance for when spending rises or falls. There are many versions. The best-known comes from Jonathan Guyton and William Klinger (2006), and here's one particular implementation, the one Morningstar tested: spending rises with inflation each year. If the withdrawal rate drifts more than 20% above where it started, the next planned withdrawal is cut 10%. If it drifts 20% below, the next one is raised 10%.

For our couple, 20% above 3.7% is about 4.4%. After the bad year, the scheduled $56,650 would be 4.7% of $1.2 million, so the guardrail triggers. Their next planned withdrawal drops 10%, from $56,650 to about $51,000. Holding Social Security level for simplicity, their total income is about $116,000 instead of $121,650, and the cut comes out of the discretionary budget.

The research case for flexibility is real. Morningstar found guardrails supported a 5.2% starting rate in a 40/60 portfolio, versus 3.9% for fixed spending. Vanguard's version caps changes with a ceiling and floor, such as no more than a 5% raise or a 2.5% cut a year.

The cost is variable income. In Morningstar's words, "there's no free lunch": higher starting spending comes with more volatile spending. A 10% cut is easy on paper and harder the year the roof needs replacing. Guardrails work best when part of the budget is truly optional and steady income covers most essentials, as it does here.

Guardrails are a withdrawal discipline, not an investment strategy. They can sit on top of a rebalanced portfolio or a bucket structure.

3. Buckets: organizing money by when it's needed

Buckets divide money by time horizon: cash for the next year or two, bonds for the next several, stocks for the long run. Spending comes from the near-term bucket, which is refilled over time.

Our couple might keep two years of their gap, $110,000, in cash, with the rest in bonds and stocks so the overall mix is still about 50/50. In the bad year they spend from cash and leave stocks alone. If stocks are still down when the cash runs low, they refill from bonds and wait.

The appeal is that you can see where the next two years of spending is coming from, which makes it easier to leave the long-term money alone.

The research is more skeptical about the finances. Michael Kitces has shown that, once the buckets are rebalanced, a bucket portfolio's allocation is nearly the same as a total-return portfolio's. When the overall allocations and the withdrawal and rebalancing rules are comparable, the economic exposure is similar; the labels don't change the risk. In a study of 21 countries over 115 years, Javier Estrada found that simple rebalanced allocations generally outperformed bucket strategies. His explanation: rebalancing systematically buys assets after they've fallen, and buckets generally don't. Money parked in cash for years also gives up growth it might have earned elsewhere.

Where buckets and total-return portfolios do differ is in the details: how much sits in cash, when and how the buckets are refilled, and how far the overall mix is allowed to drift. If cash is spent and not replenished, for example, the portfolio drifts toward stocks. Those choices can produce different outcomes, for better or worse.

Heritage uses a two-bucket version: the next few years of spending held so it doesn't depend on a good market, and the rest invested for growth, refilled when it makes sense rather than on a calendar, alongside spending guardrails. Kitces makes the fair counterpoint that a structure people can stick with may be worth more than a slightly better one they abandon. That's the reason to use buckets. It isn't extra return.

4. A guaranteed-income floor

This approach covers essentials with income that doesn't depend on markets, and invests the rest for growth and discretionary spending. Our couple's essentials are $90,000. Social Security covers $65,000, leaving $25,000. Suppose they buy a single-premium immediate annuity at the start of retirement, before any decline: a lump sum to an insurer in exchange for lifetime income. (Delaying Social Security, for those who haven't claimed, is another way to raise guaranteed, inflation-adjusted income.)

For scale, Morningstar cites a Cannex estimate that a 67-year-old woman buying a life-only immediate annuity in October 2025 would receive about 7.7% of the premium a year (about 5.8% with 3% annual increases). At 7.7%, $25,000 a year would cost about $325,000, leaving about $1.175 million invested. That payout is for a single-life contract on one person, not income guaranteed for both spouses. Joint-life income for the couple would pay a lower rate, so the same $25,000 would cost more, and inflation or refund features would raise the cost further.

If the bad year then arrives, the essentials are still covered by Social Security and the annuity. The invested portfolio will be smaller than $1.175 million by an amount that depends on how it's invested, but it now has to fund only the $30,000 of discretionary spending rather than the full $55,000 gap. That gives the couple more room to wait out a decline.

The tradeoffs are real and specific:

• Liquidity and legacy. The premium is no longer an account you can draw on or leave to heirs, unless you choose a refund or period-certain option, which lowers the payout. Then again, money you're planning to spend over your lifetime is less liquid in practice than its balance suggests.
• Inflation. A level payment loses purchasing power every year. Built-in increases help, but they're fixed at purchase, and they start you at a lower income.
• Survivor protection. Whether the income continues for a spouse depends on the option chosen.
• Insurer strength. The guarantee depends on the insurer's claims-paying ability, as the SEC's Investor.gov points out. State guaranty associations provide a limited backstop if an insurer fails, and coverage limits vary by state.
• Costs and complexity. A simple immediate annuity's cost is built into its payout rate. Variable and indexed annuities with riders can carry layers of fees and surrender charges. Some contracts are needlessly complicated and deserve skepticism, which is a judgment about specific products, not the category.

What a floor buys is insurance against a long life and bad markets, and for some people a paycheck that's easier to live with. As Vanguard notes, an income annuity works much like Social Security. Whether that insurance is worth its price depends on health, family longevity and how much you value certainty over flexibility.

Choosing among them

These approaches aren't mutually exclusive. Many good plans combine a rebalanced portfolio, a guardrail spending rule, a cash reserve and some guaranteed income. What to emphasize depends on a few questions:

• How flexible is your spending? If part of the budget can bend, guardrails are a strong tool. If almost everything is fixed, a guaranteed floor matters more.
• How much is already guaranteed? If Social Security and a pension cover the essentials, an annuity adds less.
• How much risk can you carry, financially and emotionally? The best allocation is one you won't abandon in a bad year.
• What matters most? The largest estate, the most lifetime spending, or the fewest sleepless nights each point to a different mix.

If you'd like to talk through which combination fits your own situation, you can reach me here.

This article is for general education only and isn't individualized investment, tax or insurance advice. The couple and the market decline are hypothetical, and no return or outcome is projected or guaranteed. Research figures reflect each study's own assumptions.