Inflation Risk in Retirement: How Purchasing Power Erodes, Why Retirees Feel It More, and Where the COLA Gap Comes From

The Risk That Never Shows Up on a Statement

Every other retirement risk announces itself. A bear market produces a red number on a quarterly statement. A long-term care event produces a bill. A tax change produces a letter from the IRS. Inflation produces nothing at all. The account balance is the same on Tuesday as it was on Monday, the pension check is the same as last month, and the grocery total is a few dollars higher than it was, which is not the kind of thing anyone writes down.

That silence is why inflation does more damage over a thirty-year retirement than the risks that get most of the attention. A 20% market drop is frightening, but markets have recovered from every one of them. A 20% loss of purchasing power is permanent, arrives without a headline, and compounds on itself for as long as the retiree lives.

This article puts numbers on it. It follows one household through thirty years at a few different inflation rates, shows why the retiree's basket runs hotter than the number on the news, compares how a fixed pension, a Social Security check, and a portfolio each age under the same inflation, and finishes with what actually hedges the risk and what only looks like it does.

What Inflation Risk Actually Is

Inflation risk is the chance that the income a retiree has arranged buys less than the retiree expected it to. It has three properties that distinguish it from every other risk in a plan.

It requires no bad year. Sequence-of-returns risk needs a crash, and a badly timed one. Longevity risk needs a long life. Inflation needs nothing to go wrong. A perfectly ordinary economy running at its target rate erodes a fixed income by a quarter in the first decade.

It compounds. Three percent sounds like a rounding error. Three percent a year for thirty years is a 143% increase in the cost of living. The mechanism is the same one that makes a retirement portfolio grow, running in the other direction against every dollar of fixed income.

It is invisible in nominal accounting. A retiree who spends $6,000 a month in year one and $6,000 a month in year fifteen has cut their standard of living by more than a third without ever seeing a smaller number anywhere. The damage shows up as a slowly shrinking list of things the money covers, and most people attribute that to the world getting more expensive rather than to a plan that never accounted for it.

The Working Example

For the rest of this article we will follow one household so every figure can be checked. A couple retires at 65 spending $6,000 a month, or $72,000 a year, in today's dollars. Their plan runs to age 95. General inflation is 3%, which is close to the long-run U.S. average and the rate most planning software defaults to.

Here is what it costs to keep buying exactly what $6,000 buys today:

Two figures in that table deserve to be read slowly.

The first is the last row of the third column. By 95, the same lifestyle costs $14,564 a month. A couple who planned around $6,000 and never revisited it is living on 41 cents of every dollar they intended to have.

The second is the last row of the final column. A flat $6,000 budget over thirty years totals $2,160,000. Keeping pace with 3% inflation totals $3,528,193. The difference, $1,368,193, is the price of inflation for this one household over one retirement. It is larger than the portfolio many households retire with, and it is the number no statement ever displays.

Why the Rate You Assume Matters So Much

Three percent is an assumption, and a plan is only as good as its assumptions. Here is the same household at five different inflation rates:

The spread between the first and last rows is the entire argument for not trusting a single number. At 2%, this couple needs $10,868 a month at 95. At 4%, they need $19,460. The difference between a plan built on the Fed's target and one built on the 1990s average is $8,592 a month at the end of life, and an extra $1,220,320 over the retirement. Nobody knows which of those five rows the next thirty years will look like. The 2010s ran below the first row. The three years from 2021 through 2023 ran above the last one.

That is why a plan should be tested at more than one rate rather than optimized around one. The Monte Carlo article showed that moving assumed inflation from 3% to 4% dropped a representative plan's success rate from 90.4% to 74.8%, a bigger swing than a full percentage point of stock returns. Inflation is the assumption a plan is most sensitive to and the one least often questioned.

The Retiree's Basket Is Not the CPI Basket

The headline inflation number is the Consumer Price Index for All Urban Consumers, which measures a basket weighted toward how the whole population spends. Retirees do not spend like the whole population. They spend less on transportation, education, and clothing and considerably more on healthcare and housing, and the things retirees buy more of have historically inflated faster than the things they buy less of.

The Bureau of Labor Statistics has tracked an experimental index for households headed by someone 62 or older, the CPI-E, since the early 1980s. Over its history it has run about two-tenths of a percentage point a year above the CPI-W, the index Social Security uses for its cost-of-living adjustment. That sounds small. Compounded over thirty years it is another 6% of purchasing power gone.

Healthcare is the reason. The standard Medicare Part B premium is $202.90 a month in 2026. Ten years earlier it was $121.80. That is a two-thirds increase in a decade, roughly 5% a year, against general inflation that averaged well under 3% over the same period. Prescription drugs, supplemental coverage, dental, hearing, and long-term care have followed similar paths.

Here is our household's $6,000 budget broken into categories, with healthcare inflating at one and a half times the general rate, leisure and travel at eight-tenths of it, and food slightly above it:

The category-weighted budget at 95 is $15,715 a month rather than the $14,564 a flat 3% produced. The extra $1,151 a month is almost entirely healthcare. Healthcare was 15% of this couple's spending at 65 and is 21% of it at 95, and that is before a single long-term care event. See long-term care planning for what one of those does to the picture.

Three Income Sources, Three Different Exposures

A retirement income plan is usually built from three kinds of money: Social Security, a pension or annuity, and a portfolio. Inflation treats each one differently, and the differences are large enough that two households with identical total income in year one can be in very different positions by year twenty.

The fixed pension

A traditional pension with no cost-of-living adjustment is the purest form of inflation exposure there is. It is reliable, it is guaranteed, and every check buys less than the one before it. Most private-sector pensions have no COLA at all. Many public pensions have one, but it is often capped at 1% to 3%, applied only to part of the benefit, or subject to legislative discretion.

A $3,000 monthly pension with no COLA still pays $3,000 at 95. Here is what that $3,000 actually buys, in today's dollars, at 3% inflation:

The retiree who chose the pension for its safety has, by 85, lost 45% of the income they thought they had locked in. Nothing went wrong. The pension did exactly what it promised. It is the promise that was in the wrong units.

This is also the reason the pension maximization decision and the pension versus lump sum decision are so sensitive to inflation assumptions. A lump sum invested in a growing portfolio and a fixed annuity look very different at 3% inflation than at 2%.

The Social Security check

Social Security is the only inflation-indexed lifetime annuity most Americans will ever own, and it is worth understanding exactly how good the indexing is.

Benefits are adjusted every January by the change in the CPI-W over the prior year's third quarter. Since 2000 that adjustment has averaged about 2.6% a year. It has been as high as 8.7%, for 2023, and zero three times, in 2010, 2011, and 2016. The 2026 adjustment was 2.8%. The 2027 adjustment will be announced in October 2026.

The protection is real. It is also imperfect, in two ways.

First, the adjustment tracks a worker's basket rather than a retiree's, so it captures general inflation but not the retiree's excess exposure to healthcare. Second, the Part B premium is deducted from the check before it arrives, and Part B has been rising faster than the COLA. A retiree can receive a 2.8% COLA and see a smaller net increase in the deposit, or in a low-COLA year, none at all.

Here is what that gap does over time. Our couple's $2,400 monthly benefit grows at a 2.5% COLA, roughly its long-run average. Their costs grow at 3%.

Over twenty years the accumulated gap is $43,005. Over thirty it is $115,270. That is real money the portfolio has to supply, and it is the gap most plans quietly bury by inflating spending and Social Security at different rates without saying so.

But look at the last column against the pension table above. The Social Security check that started at $2,400 still buys $2,074 worth of goods at 95, an 86% hold on its purchasing power. The pension that started at $3,000 buys $1,236, a 41% hold. Half a percentage point of COLA shortfall is a nuisance. No COLA at all is a structural problem.

The portfolio

A portfolio has no explicit inflation adjustment, but it has something the other two sources lack: assets whose prices and earnings tend to rise with the general price level over time. Stocks are claims on businesses that raise their prices when their costs rise. Real estate is a claim on rents that reset. Over any thirty-year period in U.S. history, a diversified stock portfolio has outrun inflation by a wide margin.

The catch is the word over time. In any given year, stocks can fall while prices rise, and the years when both happen at once are the worst years a retiree can have. That is the intersection of inflation risk and sequence risk, and it deserves its own section.

Bonds are the more complicated case. A conventional Treasury bond pays a fixed coupon and returns a fixed principal, which makes it a fixed pension in miniature: safe in nominal terms, exposed in real terms. A bond bought at 2% when inflation turns out to be 4% loses purchasing power every year it is held. Treasury Inflation-Protected Securities and Series I savings bonds adjust their principal with the CPI and are the only fixed-income assets that hedge inflation by construction.

When Inflation and Sequence Risk Arrive Together

The 4% rule article explains that the rule's safe withdrawal rate was set by the single worst starting year in the historical record. That year was not 1929 or 2000 or 2008. It was 1966. The 1966 retiree faced a flat stock market for the next sixteen years combined with the highest sustained inflation in modern American history, which peaked above 13% in 1980.

The mechanism is worth understanding because it is the one scenario a diversified portfolio cannot fully absorb. In a normal bear market, a retiree can hold spending flat, draw from bonds, and let the stocks recover. In an inflationary bear market, spending is rising while the portfolio is falling, so the withdrawal rate climbs from both directions at once. Meanwhile the bonds meant to cushion the fall are losing value too, because rising inflation pushes interest rates up and bond prices down. Every part of the plan fails simultaneously.

That is what makes the 1970s the historical worst case rather than the Depression. A retiree in 1929 suffered a catastrophic market decline but experienced deflation, so their fixed withdrawals bought more each year and the portfolio's real burden fell. A retiree in 1966 had a milder market and a far worse outcome. Inflation, not the crash, is what sets the floor on safe withdrawal rates.

The good news is that this scenario is also rare, and it is the one the bucket strategy and dynamic withdrawal rules are best suited to. A retiree who can trim discretionary spending by 10% for a few years, and who holds a few years of near-term spending in inflation-protected bonds rather than conventional ones, has removed most of the sting from the worst decade in the record.

What Actually Hedges Inflation

Some of the things sold as inflation protection work. Some do not. Roughly in order of cost-effectiveness:

Delay Social Security. This is the cheapest inflation hedge available to any retiree, and most people do not think of it as one. Every year of delay past full retirement age raises the benefit by 8%, and the larger benefit is inflation-adjusted for life. Our couple's $2,400 benefit at 67 becomes $2,976 at 70, and every future COLA applies to the larger figure. For a married couple, the higher earner's delay also sets the survivor benefit, which means the inflation-protected floor for the surviving spouse. See when to claim and survivor benefits and the higher-earner delay.

Keep a real stock allocation for the long horizon. The 30-year table above is the argument. A portfolio expected to fund spending that rises 143% cannot be built from assets that return a fixed 4%. The growth bucket exists to fund the years when the cash and bond buckets are exhausted, and those are the years the cost of living is highest.

Use TIPS or I bonds for the near-term floor. The two to five years of spending a retiree holds in fixed income should be the part of the portfolio most protected from inflation, because it is the part that will be spent at whatever prices then prevail. Conventional bonds are the wrong instrument for that job. A TIPS ladder maturing into each of the next several years' spending is the right one.

Build spending flexibility into the plan. A retiree who can cut 10% of spending in a bad year has an inflation hedge that costs nothing to hold. The leisure and travel line in the category table is where that flexibility usually lives, and it is also the line that naturally shrinks with age.

Be careful what goes into the fixed floor. A fixed immediate annuity, a pension without a COLA, and a conventional bond ladder are all excellent tools for the risk of outliving money and terrible tools for the risk of inflation. They belong in a plan, but they should not be the whole plan. The annuity laddering approach exists partly to address this: buying guaranteed income in tranches over time rather than all at once means later purchases are priced off later, higher rates and cover later, higher costs.

Own the house. A paid-off home is an inflation hedge on the single largest line of most retirement budgets. Housing is 30% of our couple's spending, and the portion of it that is a fixed mortgage payment does not inflate at all, while the portion that is property tax, insurance, and maintenance does. A retiree who owns outright has converted the largest budget category into something closer to a real asset.

Common Mistakes With Inflation in Retirement Plans

Planning in nominal dollars. The single most common error. A retiree sets a $6,000 monthly budget at 65 and manages to it for a decade, congratulating themselves on discipline while their standard of living falls 26%. Set the budget in real dollars and let the nominal number rise every year.

Using one inflation rate for everything. Applying 3% to a budget that is 15% healthcare understates late-life costs by more than $1,000 a month in the example above. Inflate healthcare separately.

Treating the COLA as full protection. It is most of the protection, which is more than any other income source offers. It is not all of it, and the Part B deduction erodes it further.

Confusing safe with inflation-safe. A fixed pension, a fixed annuity, and a CD ladder are all safe in the sense that the nominal payment will arrive. None is safe in the sense that matters, which is whether the payment still buys the same groceries in 2046.

Stress-testing the market and not the price level. Almost every plan is run through a bear market scenario. Far fewer are run through a decade of 4% inflation, which historically has done more damage to more retirees.

Forgetting that the second spouse lives longer. For a couple, the planning horizon is the second death, not the first. Inflation exposure runs for the longer of the two lives, and it is usually the surviving spouse, on a reduced household income, who faces the highest prices. See the widow's penalty for the tax half of that problem.

Why This Matters for Advisors

Inflation is the easiest risk to leave out of a client conversation and the hardest one to explain after the fact. A market decline generates a phone call and a chance to reassure. Inflation generates a slow, unattributed decline in the client's satisfaction with a plan that is doing exactly what it was designed to do.

The conversation that works is the one built around the first table in this article. Show the client what their own budget costs at 75, 85, and 95. Most have never seen it, and the year-30 figure is usually a genuine shock. Then show them which of their income sources will keep up and which will not, in the same table, side by side. A client who sees that their pension buys $1,236 at 95 while their Social Security check buys $2,074 understands the case for delaying Social Security without a single word about break-even ages.

It also reframes the allocation conversation. A client who wants to move to all bonds at 65 for safety is asking to hold the one asset class with no inflation protection for a thirty-year period in which prices will more than double. Putting the two tables next to each other makes that visible in a way a risk-tolerance questionnaire never will.

And it belongs in the file. An inflation assumption is a recommendation. If a plan's success depends on 2.5% inflation and Social Security keeping pace with it, that assumption should be stated, documented, and revisited every year, because it is the single input a client cannot see and the one most likely to be wrong.

How RetirementForge Helps

The Inflation Impact Calculator produced every figure in this article. It projects a household's budget year by year at any inflation rate, breaks it into categories with their own multipliers so healthcare can run hotter than the rest, and shows the cumulative cost of inflation over the plan alongside the purchasing power lost at retirement and at the end of the horizon. It is built for the conversation described above: a client's own numbers, at 75, 85, and 95, on one screen.

Pair it with the Social Security Optimizer to model the COLA-adjusted benefit at each claiming age and see how much inflation-protected income a delay buys, the Monte Carlo Analyzer to test a plan across a range of inflation assumptions rather than one, and the Income Gap Analyzer to see which income sources keep pace and which fall behind. Every assumption is documented and every client session is captured in an immutable audit trail. Get started free and show a client what their budget costs at 95 in about a minute.

This article is for educational purposes only and does not constitute financial, tax, or legal advice. All figures are outputs of a projection run against stated assumptions and are illustrative, not predictive. Historical inflation, Social Security cost-of-living adjustments, and Medicare premiums are drawn from published government figures and are subject to revision. Consult a qualified financial advisor for guidance specific to your situation.