THE WALL· Legal← Back to dashboard

Disclaimers

Last updated: 2026-06-03

We try to be honest about what we know, what we don’t, and how each piece of data was produced. Every panel on the dashboard that uses an estimate, a proxy, or AI-assisted text is labeled inline. This page explains what the labels mean and states the product’s overall posture in one place.

1. Informational and educational only

The Wall is not personalized financial, tax, legal, or accounting advice. Nothing on this site is a recommendation to buy, sell, hold, or otherwise transact in any security, asset, or financial product. We are not a registered investment adviser, broker-dealer, or fiduciary, and we are not acting in any such capacity by publishing this dashboard. You are solely responsible for your own decisions and should consult a qualified professional before acting on anything you see here.

2. Observatory, not advisor

We help you see what’s moving in markets, Congress, the economy, housing, and the AI race. We do not tell you what to do about it. We do not publish price targets, model portfolios, or buy/sell signals. We do not score any specific ticker as a recommendation.

3. Delayed disclosures are not real-time signals

Several sections of the dashboard surface data that is disclosed after the underlying event happens:

  • Congressional trades. Members of Congress are required to file Periodic Transaction Reports (PTRs) within 30 to 45 days of a trade, and disclosures are sometimes filed late. What you see on the Congress tab is at least weeks behind when the trade actually executed.
  • Capital / Money Movers filings. Form 4 (insider) and 13F (institutional) filings are disclosed days to weeks after the position was taken. 13F is a quarterly snapshot, not a live ledger.
  • Economic data releases. Macro indicators (CPI, jobs, GDP) are reported on official cadences with revisions; the first print is rarely the final number.

These data are useful for context and accountability. They are not useful for trading alongside the subject in real time, and we strongly discourage anyone from treating them that way.

4. Proxy and composite measures

Some questions don’t have a single ground-truth dataset. Where that’s the case, we either build a composite from several reasonable inputs or use a proxy — a directional stand-in for the unmeasurable thing. Both are clearly labeled inline:

  • COMPOSITE badge: a single number assembled from multiple weighted inputs. Hover the badge to see which inputs feed the score.
  • PROXY badge: this isn’t the thing itself, just our best public stand-in. Use directionally.

Examples: the Market Mood Dial is a 5-pillar composite. The housing affordability score combines FHFA HPI, ACS income, and mortgage rates into a composite, anchored to NAR’s published methodology. AI Race “value” rankings are a composite of benchmark score and per-token price.

5. AI-assisted and AI-generated content

Some text on the dashboard is written or shaped by a language model. We distinguish three states:

  • CURATED: An editor hand-picked this from primary sources. No model was in the loop. Stories, hero items, and highlighted disclosures fall here.
  • AI-ASSISTED: A model helped draft the text; an editor reviewed it before publishing. Numbers come from the linked source — the model handled wording, not facts.
  • AI-GENERATED: A model produced this from underlying source data with minimal review. Treat as a directional read, not a verified statement. The source link is always available; click it before acting.

Language models can hallucinate, miss nuance, or compress out important caveats. We do our best to keep this kind of content narrow and to favor the source over the summary.

6. AI Race specifics

  • We rank models on third-party public benchmarks (SWE-Bench Verified, MMLU, GPQA, LiveCodeBench, and others). We do not rank by lab self-reports unless explicitly tagged with the LAB-REPORTED evidence tier.
  • Each capability lane shows an evidence tier: verified ≥ independent ≥ lab-reported.
  • A model may appear in our dataset based on a lab-reported score before independent verification exists. We flag this explicitly. Never confuse a lab’s announcement with a replicated result.

7. Housing specifics

  • Affordability scores combine FHFA HPI (price index), Census ACS B19013 (income) and B25077 (home value), and Freddie Mac PMMS (mortgage rate) via the NAR HAI methodology. The score is a composite — useful for ranking, not for individual housing decisions.
  • Migration data is from the IRS Statistics of Income (SOI) county-to-county migration, with an inherent lag of roughly two tax years.
  • We do not use Zillow, Redfin, Case-Shiller, CoreLogic, or ATTOM data anywhere on the site (commercial licensing constraints).
  • Auction Watch aggregates public-record property-sale notices (sheriff execution sales, federal agency sales and REO inventories) and enriches them with county parcel records, code-enforcement cases and FEMA flood layers. It is an index of public records, not an offer, solicitation, listing service, or recommendation to buy any property.
  • Auction and distressed properties are typically sold as-is. We do not research title, liens, occupancy, redemption rights, upset-bid status, or interior condition. Notices can be postponed, withdrawn, or amended after we ingest them, and our copy can lag the official source. Always verify directly with the county clerk, sheriff, or federal agency — and consult an attorney — before bidding on or purchasing any property.
  • The Auction Watch Opportunity Score is a rule-based composite of the listed public-record inputs (bid vs. assessed value, time to sale, code-enforcement cases, flood zone, metro price momentum, data completeness). Higher risk inputs (such as a FEMA high-risk flood zone) lower the score. It ranks which records may merit attention; it is not a valuation, an estimate of market value, or investment advice. County assessed values are not market prices.
  • Auction Watch surfaces parcels, not people: we do not display owner or defendant names even where they appear in the underlying public records.

8. Markets specifics

  • Market data comes from Tiingo (equities and ETFs) and FRED (yields, commodities, volatility). It may be delayed by the provider; we are not a real-time market data vendor.
  • Sector composition is a static config (top GICS members by approximate index weight) refreshed periodically, not pulled live from ETF holdings.

9. Past performance does not predict future results

Any historical chart, return number, or trend line you see on the site reflects what already happened. Markets, models, legislation, and housing dynamics all change. Nothing on this site is a forecast.

10. Errors and corrections

We will get things wrong sometimes. If you spot an error, please email hello@thewall.dashboard and we’ll investigate and correct it. Material corrections will be noted on the relevant tab.

11. Full legal terms

See our Terms of Use and Privacy Policy for the complete agreement.