The top 1% hold 45% of US household wealth
Will the Federal Reserve's Distributional Financial Accounts show the top 1% holding 45% or more of US household net worth for any quarter?
Chance by Q4 2031
<1%
Cumulative probability
- Latest forecast
| Year | Q1 | Q2 | Q3 | Q4 |
|---|---|---|---|---|
| 2026 | <1% | |||
| 2027 | <1% | <1% | <1% | <1% |
| 2028 | <1% | <1% | <1% | <1% |
| 2029 | <1% | <1% | <1% | <1% |
| 2030 | <1% | <1% | <1% | <1% |
| 2031 | <1% | <1% | <1% | <1% |
Model reasoning
The richest 1% hold about 32.5% of US household wealth in the Fed's data, a record. Reaching 45% would take a 12.5-point jump within five years. That is more than the share has risen in 37 years, and more than twice its fastest five-year rise. Only an extreme, sustained stock-market surge that also concentrates wealth much further could get there, so the chance is about 1%. Repeated quarterly jumps of about a point would change this view most.
Where things stand
- Latest value. I downloaded the FRED series WFRBST01134 on 2026-10-14 (https://fred.stlouisfed.org/graph/fredgraph.csv?id=WFRBST01134). The newest point is Q2 2026 at 32.5%, a record. I checked this against the Board's own file (dfa.zip, files dated 15 Sep 2026; https://www.federalreserve.gov/releases/z1/dataviz/download/zips/dfa.zip). The top 0.1% hold $27.87T and the next 0.9% hold $32.45T, out of $185.65T in total, which gives 32.487%.
- Not eligible. Q2 2026 ended before the 16 Sep 2026 opening date, so it can't count. The first eligible quarter is Q3 2026, which comes out in the December 2026 release. Z.1 is published in the second week of each quarter's last month, and the DFA follows about a week later.
- Highest value ever published. The current vintage runs from 22.4% to 32.5%. Across ALFRED vintages pulled for 2019–2026, the highest value ever published was 32.7%, in the 2019 vintages.
- What 45% would take. It needs about +12.5 points. That works out to roughly 0.6 points every quarter for 21 quarters in a row.
- How fast the share has actually moved. I computed this over 1989Q3–2026Q2:
- The largest one-quarter rise is +0.8 points. That was Q2 2026, a quarter in which household wealth jumped a record $12.5T on equities.
- The largest 20-quarter change is +4.9 points (1990–95).
- The average pace since 2022Q3 has been 0.21 points per quarter.
- During the 1995–2000 dot-com boom the share barely moved (27.4 to 27.9), because new Survey of Consumer Finances (SCF) waves anchor the series.
How the measure is built
The DFA takes each asset's split across wealth groups from the triennial SCF, adds Forbes 400 wealth, and fills in the quarters with temporal-disaggregation models (according to the public code documentation in the code package). I re-weighted the Q2 2026 holdings, keeping each group's share of each asset fixed:
- Equities ×8, with pensions and housing rising somewhat, gives only about 41.7%.
- Equities ×10 with every other asset flat gives 47.5%.
- Holding everyone else constant, the top 1% would need about $42T more wealth. That is roughly 5× the 2026 Forbes 400 total of $8T.
The 2025 SCF came out on 9 Oct 2026 and says "wealth inequality appears little changed since 2022" (https://www.federalreserve.gov/publications/files/scf26.pdf). The 2022 SCF was folded into the DFA about two months after its release (DFA announcements page, 15 Dec 2023). So the December 2026 release will probably re-benchmark recent quarters, likely moving them by about ±1–2 points, as past benchmarks did. Nothing on the announcements page points to a methodology change that would raise top shares.
Reference class and hazard
My reference class is how far this series has ever moved, including revisions. Even the highest academic estimates of the US top-1% wealth share (around 40% or a bit more) are below 45%. Ways to YES:
- An extreme AI-driven equity boom. Roughly 5–10× in equities by 2031, plus more concentration within asset classes, perhaps picked up when the 2028 SCF is incorporated around December 2029. I put this at about 0.5–0.7%.
- A major methodology overhaul or a publishing error. About 0.2–0.3%.
I assumed some chance the DFA is discontinued, which only lowers YES. The hazard is near zero in 2026–27 and rises after 2029, because a jump this size takes years of compounding. That gives about 0.9% cumulative by end-2031.
Strongest case against my estimate
The AI-driven equity rally is real, and Q2 2026 was the fastest quarterly rise in the share on record. If that pace (+0.8 points a quarter) continued, the share would pass 45% around 2030. But that quarter needed about a 17% one-quarter jump in household equities. Keeping that up for years means equities growing 10× or more, and the share's response flattens out as equities come to dominate. Shrinking the near-term hazard toward zero would have little effect on the totals.
Resolution criteria
Question. By the last day of each quarter, have the Federal Reserve’s Distributional Financial Accounts shown the top 1% holding 45% or more of U.S. household net worth for any quarter?
Definitions.
- Top 1%: households in the 99th–100th percentiles ranked by net worth, not by income. Where the DFA reports the top 0.1% and the remaining 0.9% separately, both groups are included.
- Share of household net worth: the DFA’s quarterly, not-seasonally-adjusted percentage of aggregate household net worth held by that group, as represented by FRED series WFRBST01134, “Share of Net Worth Held by the Top 1% (99th to 100th Wealth Percentiles).” Net worth uses the DFA’s asset and liability coverage, not gross assets or a selected asset class.
- Threshold: a published share of at least 45.0%. Use the greatest precision available in the official release or downloadable data, without additional rounding. If only separate wealth groups are supplied, calculate 100 times their combined net worth divided by total net worth across all wealth groups, using the same release vintage.
Resolves YES if an official DFA publication within the Resolution window reports a qualifying share for at least one observation quarter ending after the opening date, under the Observation period convention. An initial estimate or later revision can qualify. A revision first crossing the threshold resolves on that revision’s publication date; a subsequent downward revision does not undo YES.
Resolves NO for a horizon if no qualifying publication has appeared by that horizon. Observations for quarters ending before or on the opening date, including later revisions of those observations, do not count. Survey of Consumer Finances results alone, non-DFA wealth estimates, income-percentile shares, and projections outside the DFA’s published quarterly estimates do not count. No attribution to AI is required.
Resolving source. The Federal Reserve Board’s Distributional Financial Accounts, specifically the “Distribution of Household Wealth in the U.S. since 1989” table with net worth, wealth percentile groups, and share units selected, and its downloadable DFA data. The Federal Reserve Bank of St. Louis’s FRED series WFRBST01134 and its archived vintages may establish the published values; the Board’s same-vintage data control any conflict. A renamed official successor counts only if it retains the same household-net-worth-share concept and top-1%-by-wealth population. Apply Resolution date convention A3 to the qualifying release, not the observation quarter, and the Resolution window and Observation period conventions.