The top 1% hold 35% of US household wealth
Will the Federal Reserve's Distributional Financial Accounts show the top 1% holding 35% or more of US household net worth for any quarter?
Chance by Q4 2031
8%
Cumulative probability
- Latest forecast
| Year | Q1 | Q2 | Q3 | Q4 |
|---|---|---|---|---|
| 2026 | <1% | |||
| 2027 | <1% | <1% | <1% | <1% |
| 2028 | <1% | <1% | <1% | 1% |
| 2029 | 1% | 2% | 2% | 4% |
| 2030 | 4% | 5% | 5% | 6% |
| 2031 | 6% | 7% | 8% | 8% |
Model reasoning
The Fed's latest figure puts the richest 1% at 32.5% of household wealth. Its new 2025 wealth survey shows the very top's share falling, and folding it in, probably in December, should cut the estimate to about 29–31%. Reaching 35% by 2031 would then need a very large stock boom plus a favourable 2028 survey; I put that at about 8%. A smaller-than-expected December revision would raise this the most.
Where things stand
- Latest value: FRED series WFRBST01134 shows 32.5% for 2026 Q2. That vintage is dated 2026-09-18 and follows 31.7% in Q1. The all-time high in any vintage is 32.9% (2019 Q4, March 2020 vintage). Reaching 35.0% needs a rise of about 2.5 pp on the current basis.
- Release timing: the DFA comes out about a week after each Z.1 release, so the 2026 Q3 data should appear around mid-December 2026.
- Q3 2026 markets: the S&P 500 gained only about 2% in the quarter (CNBC, 2026-10-01). Even on the current basis, 2026 Q3 should come in around 32.6%.
The deciding factor: a downward revision is coming
- New survey out: the 2025 Survey of Consumer Finances was released on 2026-10-09 (Fed press release; report scf26.pdf).
- Past practice: the DFA folds in each new survey at its next release.
- The 2022 survey came out on 2023-10-18 and was incorporated on 2023-12-15, according to the DFA announcements page.
- Both earlier incorporations cut the top-1% share by about 1.5 pp. The ALFRED vintages show 2019 Q4 going from 32.7 to 31.2 in October 2020, and 2022 Q1 going from 31.9 to 30.4 in December 2023.
- What the new survey shows: I computed top-1% shares from the Fed's public SCF summary extracts (scfp20XXexcel.zip).
- The shares were 38.5% in 2016, 37.2% in 2019, 35.1% in 2022 and 31.3% in 2025. My 2016 figure matches the published 38.6%.
- The top 0.1% fell from 14.7% to 12.1%.
- The Fed's own report table agrees the top decile's share slipped, from 73.4% to 72.3%.
- My rough rebuild of the DFA method: the method comes from the Fed's public DFA code, which adds the Forbes 400 (worth $4.5T in 2022 and $6.6T in 2025), scales each asset class to Financial Accounts totals and re-ranks households.
- I adjusted it so 2022 matches the published 29.3%. It tracked the 2019 and 2022 survey-year values to within about 0.5 pp.
- It implies a revised 2025 Q3 value of about 28–29%, against 31.5% now.
- So the December 2026 release will probably show 2026 Q3 at roughly 29–31%, leaving a 4–6 pp gap.
Paths to YES
- Before the 2028 survey is folded in (through the September 2029 release). The share would need to rise about 5–6 pp within roughly 2.75 years. That means stocks rising from about 35% of household net worth to about 50%. This has never happened: the largest 12-quarter rise in the share's history is 4.3 pp.
- After the 2028 survey is folded in (from December 2029). This needs a reset that is higher than market moves alone would explain, plus continued stock outperformance. Two things could help: billionaire wealth is surging (the Forbes 400 reached $8T in 2026), and 2025 may have been an unusually low sample at the very top. Against that, the last three resets were all lower than market moves implied (about −1.0, −0.9 and an estimated −2.3 pp). Over 1992–2016 they were mostly positive.
Simulation
- Setup: a Monte Carlo model using stock and other net-worth growth calibrated to DFA history from 1989 to 2026.
- The share rises about 0.26 pp for each 1 pp rise in stocks' share of net worth, plus a drift fitted to the 2022–26 period.
- Revised 2025 value: 85% weight on about 28.5%, 15% weight on about 30%.
- Reset at the 2028 survey: normally distributed with mean −0.25 pp and standard deviation 1.5 pp.
- Results: a central 7.5% chance by the end of 2031, with a range of 4–11% across the alternative assumptions I tried.
- Final figure: I added a small allowance for method changes or a delayed revision, giving about 8%.
Strongest case against my estimate
My rebuild uses public rather than internal survey data and could be off by about 1 pp. If the revision turned out mild (a 2026 Q3 value near 31.5%) and the AI-driven stock boom kept its 2023–26 pace, the chance could reach about 20%. The chance of no revision at all looks under 5%.
Resolution criteria
Question. By the last day of each quarter, has the Federal Reserve's Distributional Financial Accounts shown the top 1% holding 35% or more of U.S. household net worth for any quarter?
Definitions.
- Share: FRED series WFRBST01134, “Share of Net Worth Held by the Top 1% (99th to 100th Wealth Percentiles),” measured quarterly in Percent of Aggregate, Not Seasonally Adjusted, from the Federal Reserve Board's Distributional Financial Accounts (DFA).
- Top 1%: households ranked by net worth, not income; includes both the top 0.1% and the remaining 0.9% of the top 1% where the Board displays these separately.
- Household net worth: assets minus liabilities under the DFA's household wealth definition; the denominator is aggregate household net worth in the same DFA release, not gross assets, financial assets alone, or national wealth.
- Eligible observation: a quarter ending after the opening date, under the Observation period convention; the earliest eligible quarter is 2026 Q3. FRED dates quarterly observations by the quarter's first day.
Resolves YES if an eligible quarterly observation is published at 35.0% or higher within the Resolution window and on or before the horizon. Use the officially reported share at its published precision, without further rounding to reach the threshold; ignore floating-point serialization artifacts. An initial estimate or subsequent revision qualifies. A revision crossing the threshold resolves on its publication date, and later downward revisions do not undo YES. An officially designated successor series measuring the same share counts.
Resolves NO if no qualifying publication exists by the horizon. Observations for 2026 Q2 or earlier do not count, even if revised after opening. Estimates by other organizations, income shares, shares of selected asset categories, and projections for future quarters do not count. The DFA's standard model-based estimates of completed quarters do count.
Resolving source. Federal Reserve Board, Distributional Financial Accounts, specifically the Distribution of Household Wealth table and downloadable data for net worth by wealth percentile; Federal Reserve Bank of St. Louis FRED series WFRBST01134 is an accepted official republication. The Board's reported share takes precedence over a conflicting FRED value for the same release vintage. Archived official releases or FRED/ALFRED vintages may establish what was published and when. These named statistical sources, rather than the broader Public confirmation convention A2, govern resolution. Apply the Resolution window, Observation period, and Resolution date convention A3; the relevant date is publication, not the observation quarter.