Registered 2026-07-30
1.978% – 2.925%
Recorded range · central forecast 2.452%
No coverage declared · range not scoredRead the first recorded results, compare the published forecasts, and inspect the evidence behind them.
FIRST RECORDED REVIEW · 15 OCTOBER 2026
METRIC: U.S. BOTTOM-50% SHARE OF HOUSEHOLD NET WORTH · 2026 Q2
The bottom half of U.S. households held 2.304% of household net worth in the retained Federal Reserve estimate for 2026 Q2. Both ForeGlass central forecasts were 2.452%, or 0.148 percentage points above that estimate.
In everyday terms: of every $100 of total U.S. household net wealth, the least-wealthy half of households collectively held about $2.30. This is the group’s combined share, not an amount held by each household.
Federal Reserve · 2026 Q2
Same central value in two records
pp = percentage points
THE FOREGLASS RESULT
The July range contained the reported estimate; the August range did not. Range containment and central-value error answer different questions. The interval score prices both at once: the range's width plus 2/(1 − coverage) times the distance outside it, at the coverage the record declared, lower is better; the July record declared no coverage, so it carries no score. A wider range is not, by itself, evidence of a better forecast.
1.978% – 2.925%
Recorded range · central forecast 2.452%
No coverage declared · range not scored2.339% – 2.557%
Recorded range · central forecast 2.452%
Interval score 0.565 pp at 80% coverageAbsolute error is the distance between the central forecast and the reported estimate, in percentage points. The ForeGlass relative error is 6.41% of the reported estimate; it is not a success rate. The July record stated no probability level for its range; the August record declared 80% coverage. Those are different declarations, so the ranges are not presented as equivalent confidence intervals. These coverage annotations were checked against the original records; this public projection does not expose a coverage field.
The recorded comparison anchor is 2.471%, from 2025 Q1. Its absolute error is 0.167 percentage points, compared with 0.148 for the two ForeGlass central forecasts. The anchor is preserved as recorded; it is not the same quarter one year before the 2026 Q2 target. This single comparison is not evidence of general predictive skill.
THE PUBLISHED COMPARISON
Smallest absolute error first. Ordering uses unrounded errors for this one question. Equal errors are ordered by registration date, then record ID; displayed errors are rounded. This is a comparison with one reported estimate, not an overall ranking of forecasting quality.
“With engine context” means the model received simulation information. The first-round “cold” prompt design also reused the simulation’s sealed claim, which could contain forecast numbers. The original verbatim prompts were not retained. “Cold” therefore does not establish independence from the engine for this cohort. Ensembles combine model forecasts. These input conditions matter when interpreting agreement.
All 19 published central forecasts were above the reported estimate. This describes these records on this one question, not performance on other outcomes.
| Forecaster / input | Registered | Central | Recorded range | Error ↑ | Interval score |
|---|---|---|---|---|---|
Ensemble · medianWith engine context · retiredRecord IDpred-2026-08-17-tourney-us-dfa-bottom50-networth-share-interval-ens-median-grd | 2026-08-17 | 2.450% | 2.340% – 2.560% | 0.146 pp | 0.577 pp |
Ensemble · medianCold arm · first roundRecord IDpred-2026-08-17-tourney-us-dfa-bottom50-networth-share-interval-ens-median-raw | 2026-08-17 | 2.450% | 2.340% – 2.560% | 0.146 pp | 0.577 pp |
Ensemble · trimmed meanCold arm · first roundRecord IDpred-2026-08-17-tourney-us-dfa-bottom50-networth-share-interval-ens-trimmed_mean-raw | 2026-08-17 | 2.450% | 2.342% – 2.560% | 0.146 pp | 0.599 pp |
DeepSeek-V4-ProCold arm · first roundRecord IDpred-2026-08-17-tourney-us-dfa-bottom50-networth-share-interval-foundry-deepseek-v4-pro-raw | 2026-08-17 | 2.450% | 2.340% – 2.560% | 0.146 pp | 0.577 pp |
gpt-4oWith engine context · retiredRecord IDpred-2026-08-17-tourney-us-dfa-bottom50-networth-share-interval-foundry-gpt-4o-grd | 2026-08-17 | 2.450% | 2.340% – 2.560% | 0.146 pp | 0.577 pp |
gpt-4oCold arm · first roundRecord IDpred-2026-08-17-tourney-us-dfa-bottom50-networth-share-interval-foundry-gpt-4o-raw | 2026-08-17 | 2.450% | 2.350% – 2.550% | 0.146 pp | 0.657 pp |
Grok-4.3With engine context · retiredRecord IDpred-2026-08-17-tourney-us-dfa-bottom50-networth-share-interval-foundry-grok-4-3-grd | 2026-08-17 | 2.450% | 2.340% – 2.560% | 0.146 pp | 0.577 pp |
Grok-4.3Cold arm · first roundRecord IDpred-2026-08-17-tourney-us-dfa-bottom50-networth-share-interval-foundry-grok-4-3-raw | 2026-08-17 | 2.450% | 2.340% – 2.560% | 0.146 pp | 0.577 pp |
Kimi-K2.5With engine context · retiredRecord IDpred-2026-08-17-tourney-us-dfa-bottom50-networth-share-interval-foundry-kimi-k2-5-grd | 2026-08-17 | 2.450% | 2.380% – 2.520% | 0.146 pp | 0.897 pp |
gpt-5-miniCold arm · first roundRecord IDpred-2026-08-17-tourney-us-dfa-bottom50-networth-share-interval-gpt-5-mini-raw | 2026-08-17 | 2.450% | 2.340% – 2.560% | 0.146 pp | 0.577 pp |
kimi-k3Cold arm · first roundRecord IDpred-2026-08-17-tourney-us-dfa-bottom50-networth-share-interval-kimi-kimi-k3-raw | 2026-08-17 | 2.450% | 2.340% – 2.560% | 0.146 pp | 0.577 pp |
Ensemble · trimmed meanWith engine context · retiredRecord IDpred-2026-08-17-tourney-us-dfa-bottom50-networth-share-interval-ens-trimmed_mean-grd | 2026-08-17 | 2.451% | 2.340% – 2.559% | 0.146 pp | 0.573 pp |
ForeGlassForeGlass EngineRecord IDpred-2026-07-30-025 | 2026-07-30 | 2.452% | 1.978% – 2.925% | 0.148 pp | Not scored · no coverage declared |
ForeGlassForeGlass EngineRecord IDpred-2026-08-11-069 | 2026-08-11 | 2.452% | 2.339% – 2.557% | 0.148 pp | 0.565 pp |
DeepSeek-V4-ProWith engine context · retiredRecord IDpred-2026-08-17-tourney-us-dfa-bottom50-networth-share-interval-foundry-deepseek-v4-pro-grd | 2026-08-17 | 2.452% | 2.339% – 2.557% | 0.148 pp | 0.565 pp |
Ensemble · inverse varianceWith engine context · retiredRecord IDpred-2026-08-17-tourney-us-dfa-bottom50-networth-share-interval-ens-inverse_variance-grd | 2026-08-17 | 2.453% | 2.408% – 2.497% | 0.148 pp | 1.130 pp |
Ensemble · inverse varianceCold arm · first roundRecord IDpred-2026-08-17-tourney-us-dfa-bottom50-networth-share-interval-ens-inverse_variance-raw | 2026-08-17 | 2.458% | 2.417% – 2.499% | 0.154 pp | 1.206 pp |
gpt-5-miniWith engine context · retiredRecord IDpred-2026-08-17-tourney-us-dfa-bottom50-networth-share-interval-gpt-5-mini-grd | 2026-08-17 | 2.470% | 2.320% – 2.580% | 0.166 pp | 0.417 pp |
Kimi-K2.5Cold arm · first roundRecord IDpred-2026-08-17-tourney-us-dfa-bottom50-networth-share-interval-foundry-kimi-k2-5-raw | 2026-08-17 | 2.480% | 2.400% – 2.560% | 0.176 pp | 1.117 pp |
SOURCE AND TIMING
Federal Reserve Distributional Financial Accounts ↗
The figures are reported estimates, including modeled allocations between wealth groups. The current source may contain subsequent revisions.
SCOPE OF THIS REVIEW
Of 179 forecasts due for this review, 39 outcomes were recorded and 140 remained on hold. All 39 recorded outcomes concern the same wealth-share question.
These are separate methodological holds, not a claim that every missing result is waiting for new data.
The source-admission policy was adopted on October 14, after the outcome had been inspected. Registration, evidence selection and later scoring are different steps. Neither an unchanged hash nor a newly timestamped release proves that an outcome was unknowable when a forecast was made.
Resolve each outstanding hold under an explicit evidence policy, collect more independent outcomes, and use the errors to guide model improvements. Future evaluations should specify source admission before outcome exposure. The immediate milestone is a checkable record of this result, including its limitations.
| Lanei | Claimi | Predictedi | Forecasteri | Resolvesi | Againsti | Resulti |
|---|
How a scored row was graded. Results use the retained copy of the source in our evidence pack. For the first review, the file was retrieved October 1 and used for grading October 15; the source publication date is not established. Later agency revisions do not silently replace this recorded result. The first review's comparison anchor is 2.471% from 2025 Q1, preserved as recorded, not the same quarter one year before the 2026 Q2 target. Where an anchor is available it appears as naive_value. This edition publishes no skill figure. A row marked late_by_design was graded after the scheduled date. We show the source’s publication date when documented; otherwise it is marked “publication date unknown.” A late grade alone does not explain the delay. For a late row, release_observed carries either the documented date or unknown; an absent field is a publication error, not evidence of an on-time release. Download, publication and grading dates are distinct.
How to read a row: each row represents one registered forecast. Claim describes the committed statement (or a labelled summary where the original wording is withheld); Against names its source or grading rule. Click a row, or focus it and press Enter, to inspect the available record, hash and source-series chart. Forecast contents, outcome entries and publication visibility are distinct.
Two different comparisons: for a question with a recorded outcome, the observed value and absolute error show how each forecast compared with that estimate. The LLM median and consensus board measure agreement between forecasts. Input conditions differ: retired engine-grounded entries received the simulation forecast and are not independent alternatives to it. Other prompts or inputs may produce different results.
These are the largest differences between the ForeGlass engine and the recorded LLM consensus. Distance is measured in consensus spreads (z). Agreement and disagreement describe forecasts, not their correctness. Inputs are published in ledger.json; registration dates and independent timing evidence must be checked separately.
| Question | Due | ForeGlass | LLM median | Gap | z | LLMs |
|---|
How to use this view. It identifies differences to investigate when admissible outcomes are recorded. It is not a ranking of forecasting quality. Some results or forecast contents remain unavailable at this public tier.
Agreement, not accuracy. This ranking shows agreement with the LLM consensus, not accuracy. Available individual results appear in the Result column. Agreement alone does not show that a forecast is correct.
Agreement, not accuracy. This ranking shows agreement with the LLM consensus, not accuracy. Available individual results appear in the Result column. Agreement alone does not show that a forecast is correct. This board does not compute an accuracy ranking, and the Accuracy column is not populated from outcomes.
In plain terms: a forecast about a part should be consistent with the whole it belongs to. Each chain takes an officially published total, splits it into the pieces the statistics agency itself publishes, adds them back up, and shows how far off the round trip lands. A rung marked ◆ carries one of our sealed forecasts. Each chain is a group: open the one you want. The per capita switch redraws the ladders per person or per household where the chain has a stated denominator - that view is a ratio, so the shares, the running total and the residual stay on the totals.
These are measured, published values reconciling with each other - the frame our sealed forecasts sit inside.
Where the parent total is independently compiled (the World Bank's EU aggregate), landing on it checks our member set, joins and arithmetic against an outside compiler. Where the parent is only the sum of its own published bands, exact agreement is the expected result - it verifies arithmetic, not two independent measurements.
This page brings together registered forecasts, hash commitments and admitted outcomes. Registration, source admission, scoring and public release are separate steps. The first recorded review and its limitations are explained in First results; the ledger lets you inspect the individual public records.
hit. The headline result of a range is its interval score, a proper score: the range's width plus 2 / (1 − coverage) times the distance by which the reported estimate fell outside the range, at the coverage the record declared (a record that declared none has no score). Lower is better. A narrow range that misses is penalised, a wide range that contains the estimate pays for its width, and the score is in the metric's units, so it compares records on one question and never across questions. Relative absolute error divides central error by the magnitude of the reported estimate and multiplies by 100. It is not a success rate. Where a published record has naive_value, its comparison-anchor error can also be calculated. This edition publishes no skill figure or aggregate winner. For the first review the anchor is 2025 Q1, with its original historical label preserved in the record.release_observed contains a documented publication date or the explicit value unknown for a late row; if neither is present, the page reports the missing metadata. Retrieval and grading dates must not stand in for an unknown publication date.Disclosure boundary: this public edition does not include model internals, calibrations or full run lineage. You can inspect the published forecasts and reported results; you cannot reproduce the complete simulation from this page alone.
Vendor series:
The lookup searches what this page already holds: the sealed hash of each published forecast, the held commitments and the report source’s digest in ledger.json, and the ledger file’s own digest, computed in your browser. Nothing is sent anywhere.
Alongside this page sits ledger.json — the machine-readable ledger behind everything shown here.
Download the exact report source for this releaseThe ledger identifies the report source and its SHA-256 digest. The source digest covers the report before website navigation and related videos were added; it does not cover this composed web page.
index.json. Each entry is addressed by its own date (ledger-archive/<date>.json), so a claim made months ago can be checked against the bytes that carried it rather than against today’s ledger. What the index is: a record of what the last site build saw in that directory — not a guarantee that every release is listed, and not a promise that any particular entry is still served. Neither the index nor this page republishes the underlying source data; those are linked from their own landing pages, byte-for-byte copies are not redistributed here, and a link that resolves is not a reconstruction.ledger.json carries a commitments array: one {id, hash, sealed date} row per held forecast. When a held forecast is released, its hash on this page must equal the hash committed earlier — a forecast rewritten in between cannot match.ledger.json at an independent web archive. That timestamp is outside our control, which is what makes the ordering claim checkable by someone who does not trust us.Limits of the first review. Of 39 recorded outcomes for one shared question, 19 are public results here. The other 20 are hash-only commitments. These cohorts have different registration dates and do not establish an overall winner, a general accuracy rate or long-run interval calibration. Source admission happened after outcome exposure.
Among the 18 public records that declared 80% interval coverage, none contains the reported estimate. Among the 20 hash-only records declaring 80%, 17 contain it. The remaining public record declared no coverage level; its range contains the estimate and is the widest of the 39. Median relative central error is 6.3% in the public cohort and 7.6% in the hash-only cohort. These are different descriptions of one shared outcome, not independent tests of 38 forecasters or an explanation of why the cohorts differ.
The ledger's withheld count describes an entirely excluded visibility class across the whole registry. It is separate from both the 19 public results and the 20 hash-only records for this question.
Honest limit: two of the fields inside each hash are internal run identifiers, so this ledger does not let you recompute a hash from scratch — it lets you check that a published hash never changed. Independent archive snapshots are what turn that into a time-anchored claim.