Retirement Wealth at 55-74: Real Estate Share (SCF 2025)

What the Fed's own report does not show

The Federal Reserve's October 2026 report, “Changes in U.S. Family Finances from 2022 to 2025”, breaks median and mean net worth down by the age of the reference person in 10-year groups, and asset ownership by type for all families. It does not cut asset composition by net worth band within the retirement years. That is the question a reader with $250,000 or more has: at my age and my level of wealth, how much of what people like me own is a house, a building or a business, and how much is an IRA? The generic tables that rank for “net worth by age” answer a different one, the median of everyone aged 55 to 64.

We computed it from the public Summary Extract (SCFP2025.csv) and the replicate-weights file. Everything below is our arithmetic on those files, not a published Fed table.

How we computed it, and what each number means

• Who: families whose reference person (variable AGE in the extract) is 55 to 74. The SCF “family” is the primary economic unit: the economically dominant single person or couple and those financially interdependent with them.
• Bands: net worth (assets minus debts) of $250,000 to under $1,000,000, $1,000,000 to under $5,000,000, and $5,000,000 or more, in 2025 dollars, which is how the Fed states all dollar variables in the extract. A family is placed in a band using its net worth in each of the five imputed copies.
• Weights: the extract's weight WGT is the survey weight divided by 5, so that all five imputed records can be pooled to give correct population totals. We pooled the five and used WGT.
• Sample size: the unweighted number of families (implicate 1 records). We would flag any band under 100 families; none is.
• Aggregate share: the weighted dollars in an item for all families in the band divided by the weighted net worth of all families in the band. It is dollar-weighted, so wealthier families in the band count for more. It is not a median.
• Median family share: for each family, its item divided by its net worth; we take the weighted median of those ratios. Where most families own none of an item, the median is 0.
• Owners only: the median among families that own the item.
• Items: primary residence (HOUSES, gross value, mortgage not netted); other residential real estate (ORESRE, second homes, time shares, 1-4 unit properties, gross); nonresidential real estate (NNRESRE, net of its mortgages); business equity (BUS); retirement accounts (RETQLIQ). Because house and other residential values are gross while net worth is net of debt, these shares will not sum to 100%. The CSV also has the same items as a share of total assets.
• Illiquid: our label for primary residence plus other real estate plus private business equity. A home can be sold, but not in the weeks a tender window or a distribution deadline allows, and the extract has no field for how marketable each asset is.
• Standard errors: sampling variance from the 999 bootstrap replicate weights (computed on implicate 1, as the Fed's standard-error documentation describes) plus imputation variance across the five copies multiplied by 1 + 1/5; the standard error is the square root of the sum. This is our own implementation, not the Fed's SAS or Stata routine, so expect small differences. Treat differences of under about two standard errors as noise.

Who owns real estate beyond the home

The jump from the first to the second band is the large one: the share owning other real estate roughly doubles, from 23.6% to 47.3%, and the share with a private business roughly doubles too. At $5 million and over, about half have a private business. Other real estate cannot be split into “rental” and “not rental” in the extract: the Fed's definition of other residential property includes “second homes, time shares, one- to four-family rental properties”.

How much of the wealth is illiquid

Aggregate share, so the families with the most dollars weigh most:

And the median family's own ratio, which answers “what is the typical family like?”:

The two tables differ on purpose. In the lowest band the house is two-thirds of aggregate net worth and the median family also holds most of its wealth in the house, because few have anything else. At $5 million and over, the aggregate illiquid share (45.3%) is well above the median family's, because a minority of business owners hold very large business stakes. Reading the aggregate as “what a typical family holds” would be wrong for that band, and reading the median as “where the dollars are” would be wrong too.

For the middle band, 47.3% of families own other real estate and the median owner's holding is $400,000. That is 22.2% of the owner's net worth at the median, a large position, and the dollars are concentrated in a few: other real estate is 12.4% of the band's aggregate net worth.

Does it change from 55 to 74?

Among families with at least $250,000, ownership of other real estate is 42.7% at 65 to 69 and 30.5% at 70 to 74, a drop of about three standard errors, while retirement-account holding falls from 84.9% at 55 to 59 to 66.8% at 70 to 74. The aggregate illiquid share does not move in a way the standard errors can separate from flat. With 249 to 335 sample families per age group, we would not read anything finer than that.

What a reader can do with this

• Locate yourself by net worth first, age second. If you are between $1 million and $5 million, the pattern for the 55-74 group is roughly half illiquid and about a third in retirement accounts. If your own mix is far from that, it is a prompt to look, not a verdict.
• Separate what can pay a required distribution from what cannot. Only retirement accounts owe an RMD, and an illiquid asset held inside the IRA is a special case, covered in our RMD guide for illiquid IRA assets.
• Compare the owner's choice. If you are weighing a rental against a pooled real estate investment, see real estate crowdfunding vs rental property and best passive real estate income.
• Ask what the mix would mean in your plan. Illiquid means not available on a date you choose; it can still be a sensible holding. That is a question for someone who can see your whole balance sheet.

If you want to compare what advisers charge first, see fee-only financial advisor and how much a financial advisor costs.

FAQ

Sources, read and saved on October 20, 2026: Board of Governors of the Federal Reserve System, “Changes in U.S. Family Finances from 2022 to 2025: Evidence from the Survey of Consumer Finances” (October 2026, federalreserve.gov/publications/files/scf26.pdf); Survey of Consumer Finances 2025 Summary Extract Public Data (SCFP2025, federalreserve.gov/econres/files/scfp2025excel.zip) and its SAS macro with variable definitions; the 2025 replicate weight file (scf2025rw1x.zip) and the Standard Error Documentation; the SCF home page (federalreserve.gov/econres/scfindex.htm). All shares, medians and standard errors are our calculations. This is analysis of public documents, not investment, legal or tax advice.