How Much Can a Retiree Actually Spend? A Modern Methodology for Advisors

Back to Resources

Sooner or later, every client asks the question that retirement planning exists to answer: “How much can I spend?” The honest answer is a dollar amount. It is a monthly retirement paycheck, funded from every income source the household has (Social Security, pensions, rental income, portfolio withdrawals), set at a risk level the client has chosen, and recalculated as markets, inflation, and the plan itself move. For one couple, the number might be $8,200 a month. For a widowed client with a smaller portfolio and a survivor benefit, it might be $4,800. It is never a percentage, and it is never a probability score.

Most of what retirees and advisors find when they go looking for this answer is something else. The 4% rule offers a starting withdrawal rate. The 80% replacement ratio offers a budgeting target borrowed from the savings years. Probability of success offers a grade. Each one answers a question adjacent to the one being asked. This article covers why those frameworks fall short of the spending question, how risk-based guardrails answer it directly, and what the answer looks like when an advisor delivers it to a real household.

The question behind every plan review

When a client asks how much they can spend, they are not asking for a prudent withdrawal rate. They are deciding things: whether to book the river cruise this year or keep postponing it, whether the kitchen finally gets renovated, whether they can help a grandchild with tuition without quietly worrying about it for the next decade. The question is concrete, monthly, and denominated in dollars. The answer should be too.

It also depends on the entire household balance sheet. Two households with identical $1.4 million portfolios can have very different answers if one has a defined benefit pension and rental income and the other has neither. Any methodology that looks only at the portfolio is answering a narrower question than the one the client asked.

Getting the answer wrong is costly in both directions. Running out of money is the famous failure. The common one is quieter: the household that underspent and suffered anxiety for thirty years and ended with far more money than they had hoped, having skipped trips and experiences with friends and family that cannot be rescheduled. A real answer to the spending question has to guard both sides of that ledger.

Why the standard answers fall short

The 4% rule answers a withdrawal question, not a spending question

William Bengen’s 1994 research asked what initial withdrawal rate, adjusted for inflation each year, would have survived every historical retirement period for a 60/40 stock-bond portfolio. The answer was about 4%, and it was a genuinely important finding. But the rule was built to answer “what constant withdrawal rate survived history,” and that is not the question retirees ask.

The first problem is that the rule’s famous failure case barely exists. The scenarios in which a 4% withdrawal actually depletes a portfolio are worse than we’ve ever seen historically. The realistic failure is the opposite one: a retiree who follows the rule mechanically through decades that turn out fine, spending far less than their plan could have supported the whole way. The 4% rule fails by underspending far more often than it could ever fail by depletion.

The second problem is structural. For most real plans, there is no such thing as a particular, constant withdrawal rate. Withdrawals spike in the years between retirement and Social Security, drop when benefits begin, and shift again when the mortgage is paid off or the car needs replacing. A constant rate describes a household that does not exist.

The 80% replacement ratio is a savings tool, not a spending answer

The replacement ratio says a retiree will need roughly 70% to 80% of preretirement income. As a target for the accumulation years it has a real job: it gives a 45-year-old a savings goal. But it anchors to what the client used to earn, not to what their plan can now support. A retiree does not need an estimate tied to an old salary. They need a number computed from what they actually have.

Guyton-Klinger guardrails: the right instinct, withdrawal-rate mechanics

Jonathan Guyton’s 2004 paper and the 2006 Guyton-Klinger collaboration deserve genuine credit for bringing guardrails into mainstream advisor practice: spending that adjusts within named rules instead of a fixed rate followed indefinitely. The instinct was right. The mechanics, built entirely on withdrawal rates, have not held up well in practice. Research independently produced by Fitzpatrick and Tharp, Pfau, and Jeske has shown that withdrawal-rate triggers ignore everything in a plan except the ratio of spending to portfolio, stay static as the client ages, and often force large spending cuts that turn out to have been unnecessary. (For the full critique, see why Guyton-Klinger guardrails are too risky for most retirees.) The concept of guardrails is sound. Doing it well requires the whole plan.

Probability of success answers a question no client ever asked

Probability of success is the percentage of simulated futures in which a portfolio did not run out of money. (The underlying simulation method is legitimate; the score derived from it is the problem.) To the metric, “success” means the money lasted, even if the client spent three decades skipping the things they retired to do. To the client, success means living the best life they can with the time and money they have. It is the same word carrying nearly opposite meanings.

The score is also one-sided. It only measures how careful someone is being, so it pushes every client toward more caution. A 100% probability of success is, mathematically, a 100% risk of underspending and regret.

The fix is not presenting the score with better caveats. It is removing the score from client conversations entirely and delivering the answer in dollars instead. Clients will not miss it; no client ever walked into a review meeting hoping to learn their probability of success. (For the longer treatment, see why probability of success can mislead advisors.)

A methodology that answers in dollars

Risk-based guardrails start from the entire plan: the portfolio and its allocation, fees, Social Security timing, pensions, rental income, special expenses like a wedding or a new roof, longevity, and taxes. The software models more than a thousand scenarios of returns and inflation, producing a cloud of possible futures for that specific household, and then finds the spending level that matches a chosen risk posture. The default posture is quite balanced and reasonable: 20% overspending risk and 80% underspending risk, with nine preset positions so the posture can be tuned to the client. (For the full mechanics, see Income Lab’s guardrails-based retirement planning.)

The output is the client’s retirement paycheck (or, if you prefer, “spending capacity”): a dollar amount per month. This is what a retirement spending plan should be, and it is different from a budget. A budget is a snapshot. A retirement spending plan is a process: the plan is re-run every month, and it calls for a change only when a change is worth making.

Guardrails are portfolio balances, not spending amounts

Around the paycheck sit two thresholds, and both are expressed as portfolio balances. The upper guardrail is the balance above which the plan tells the client they can spend more. It is permission to spend more, and it is the structural defense against the underspending-and-regret outcome. A plan can reach it three ways: the portfolio grew faster than expected, spending ran lower than expected, or inflation came in lower than expected. When the upper guardrail is hit, the increase takes spending all the way back to the target for the household’s chosen risk posture.

The lower guardrail is the balance at which the plan calls for slowing down. It is a yield sign, not a cliff. A typical plan closes about 10% of the gap between current spending and the target spending at that balance, then keeps monitoring. The gradualism is deliberate: a household that cuts hard on a temporary drawdown and then watches the market recover has been whipsawed for nothing. If things keep getting worse, the plan may call for another measured step. And a spending cut is not the only available response. Shifting the size or cadence of an expense (a new car every 7 years instead of every 5), revisiting Social Security claiming timing, or another plan change the advisor and client choose together can do the same work.

Between the two balances is the no-change range, the range of portfolio balances where nothing needs to change. Most months, that is where the household lives.

Monthly checks, rare changes

The plan recalculates every month. A default 5% minimum-change threshold filters out movements too small to matter, whether they come from the guardrails or from accumulated inflation; the plan calls for an inflation adjustment once inflation has cut the household’s purchasing power by 5% or more. Client-facing reviews stay at the advisor’s normal pace (quarterly, semi-annual, annual), with the monitoring running quietly in between.

The numbers themselves are point-in-time estimates, and they are supposed to move. Guardrails will almost never stay steady, and they shouldn’t. A guardrail that never moved would mean the plan had stopped paying attention.

What funds the retirement paycheck

The retirement paycheck is the household’s total monthly spending, funded from every source together. For one household, that means Social Security checks, a defined benefit pension, the rent from a duplex, and portfolio withdrawals from an IRA. The number presented to the client is never a portfolio-only figure. A client does not live on their portfolio; they live on their income, and the plan’s job is to orchestrate all of it. The funding mix underneath the number will change over time even when the number itself holds steady, and the sources, especially the accounts being drawn from, will look different in year twelve than in year one.

The shape of spending changes too. Households do not spend flat through retirement; spending tends to run higher in the active early years and lower later on, a pattern David Blanchett documented in his retirement spending research. By acknowledging we won’t spend as much in our 80s and 90s, we can spend a lot more in our 60s and 70s. The methodology models that shape directly, with age-based, flat, or custom income paths.

A worked example: Margaret and David

Margaret and David are both 65 and retiring in January. They have $1.4 million, most of it in David’s 401(k) with the rest in a joint taxable account, and no defined benefit pensions. Their claiming plan has Margaret starting Social Security at 67 and David at 70, projected at $58,000 a year combined once both benefits are in pay. Their first question for the advisor is the obvious one: “How much can we spend, starting in January?”

At the default risk posture, the plan’s answer is $8,200 a month. The advisor explains how it will be funded: for the first two years, almost entirely from withdrawals from the 401(k); when Margaret’s benefit starts, the withdrawals shrink; when David’s starts at 70, they shrink again. The paycheck holds while the sources underneath it shift. The heavy early withdrawals are planned for, not a warning sign.

Then the thresholds. “Your upper guardrail is $1.7 million. If your portfolio reached that point, the plan would tell us you can spend more, and we would take your spending all the way back up to your target. Your lower guardrail is $1 million. If your portfolio fell to that point, we would slow down, moving toward roughly $6,900 a month in measured steps rather than dropping there overnight. A spending cut wouldn’t be the only option, either; we could also look at the timing of your Social Security or the cadence of your bigger expenses. Between those two balances, nothing changes, and most months that is exactly where you’ll be.”

When Margaret asks the question underneath the question, the one about what happens if they retire into another 2008, the advisor does not reach for the math. Income Lab’s Retirement Stress Test runs their exact plan through real historical sequences, including the Global Financial Crisis, 1970s stagflation, and the Great Depression, and shows when the guardrails would have been hit, what the adjustments would have looked like, and how the plan came through. Walking a client through two or three of those periods does more for their understanding than an hour of methodology ever will.

For the meeting structure around all of this, see how to present a guardrails-based retirement plan to clients.

Two more households, two different answers

Elena is 72 and widowed, with $580,000 and a $2,400 monthly survivor benefit from the Social Security Administration. Her retirement paycheck at the default posture is $4,800 a month: the survivor benefit plus about $2,400 in portfolio withdrawals. Her lower guardrail is $430,000. If her portfolio fell to it, the plan would call for moving toward $4,300 a month in measured steps, and the entire adjustment would land on the portfolio side, because the survivor benefit keeps arriving no matter what markets do.

Robert and Carolyn, both 67, have $2.4 million, an $1,800 monthly defined benefit pension, and both Social Security benefits in pay. Their retirement paycheck is $11,200 a month. Their lower guardrail is $1.75 million, with spending moving toward $10,100 a month if it is ever hit.

Three households produce three different dollar answers from one methodology. A withdrawal-rate rule would have handed all three the same percentage and called it an answer.

To see the methodology run on a household like one of these, Book a walkthrough.

How Income Lab delivers the number

Income Lab calculates the retirement paycheck and both guardrails for every plan, and tracked plans update automatically every month: balances refresh, plan duration adjusts as clients age, inflation accumulates, and the guardrails are tested. The advisor is notified when a plan is calling for a change and otherwise left alone. The plan stays current, including inflation and cost-of-living adjustments, without any extra work from the advisor.

The numbers display in today’s dollars or future dollars, whichever makes the conversation with the client clearer.

Penny, Income Lab’s AI for retirement income planning, can draft the client-ready summary of the paycheck and the guardrails at the advisor’s request. The advisor reviews, adjusts, and signs off before anything reaches the client.

FAQ

What is the 4% rule and does it still work?

The 4% rule, introduced by William Bengen in 1994 using a 60/40 stock-bond portfolio, says a retiree can withdraw 4% of starting principal in the first year and adjust that dollar amount for inflation every year after. In nearly every historical scenario it preserved capital; the cases where it depletes a portfolio are worse than we’ve ever seen historically. Its practical failure runs the other way: retirees who follow it mechanically through ordinary markets badly underspend what their plans could have supported. It remains a useful research benchmark. It is not something a real household should implement, because it answers “what constant rate survived history” rather than “how much can this household spend right now.”

How do I calculate sustainable retirement spending?

Start from the full plan rather than the portfolio alone: investments, Social Security and pension timing, other income, special expenses, longevity, and taxes. Model many scenarios of returns and inflation, choose a risk posture, and find the monthly dollar amount that matches it. Then recalculate every month and change the number only when a guardrail is hit or accumulated inflation crosses the minimum-change threshold (5% by default). The output is a spending amount in dollars, not a withdrawal rate and not a score.

What is a retirement paycheck?

The retirement paycheck is the monthly dollar amount a retirement plan can sustain at a chosen risk posture, funded from all of the household’s income sources together: Social Security, pensions, rental income, and portfolio withdrawals. It is the direct answer to “how much can I spend?”, delivered in the units clients actually think in. The number updates as the plan and markets move, and it should.

How much should I spend each month in retirement?

There is no general answer, only a household-specific one. A rule of thumb like the 4% rule implies roughly $3,300 a month per $1 million of portfolio in the first year, but that figure ignores Social Security, pensions, the shape of spending over a lifetime, and everything markets do after year one. The three households in this article, with portfolios from $580,000 to $2.4 million, landed at $4,800, $8,200, and $11,200 a month. The honest version of the answer is a number computed from the full plan and recalculated as conditions change.

Is the 80% replacement ratio accurate?

It is accurate at the job it was designed for: giving a working-age saver a rough target for how much income their savings will eventually need to replace. It is not a retirement spending answer. It anchors to preretirement salary instead of to the actual portfolio, income sources, and plan the retiree now has. Use replacement ratios to set savings targets. Use a plan-based methodology to set spending.

How do you adjust retirement spending over time?

Monitor the plan monthly and change spending only when a change is worth making. In a risk-based guardrails plan, that means an adjustment when the portfolio crosses a guardrail balance or when inflation has cut purchasing power by 5% or more (the 5% threshold is a default). Upward adjustments go all the way back to the target for the chosen risk posture. Downward adjustments are gradual, closing about 10% of the gap at a time so a temporary drawdown doesn’t whipsaw the household. Client conversations stay at your normal pace (quarterly, semi-annual, annual), plus any conversation a guardrail triggers in between.

Recommended reading

• The complete guide to retirement income guardrails
• How to present a guardrails-based retirement plan to clients
• Why Guyton-Klinger guardrails are too risky for most retirees
• Derek Tharp and Justin Fitzpatrick’s risk-based guardrails research
• Why probability of success can mislead advisors
• Penny, Income Lab’s AI for retirement income planning

Every framework in this article was built by people trying to make retirement income safer. Only one of them answers the question the client is actually asking: “How much can I spend?” The 4% rule answers with a rate. The replacement ratio answers with a savings target. Probability of success answers with a grade. Risk-based guardrails answer with a retirement paycheck in dollars, the portfolio balances that would change it, and the knowledge that most months the answer is no change at all. Book a walkthrough to see the answer calculated on a household like the one in your next meeting.