Full record

Missmarkets

The S&P 500 falls about 4-5% within roughly a month of late August 2017.

A scored prediction by Tom Lee, made on .

TweetQuoted directly by a news outletmarkets

Scored prediction

Missmarkets

The S&P 500 is set for about a 4 percent to 5 percent decline

The S&P 500 falls about 4-5% within roughly a month of late August 2017.

Made on
2017-08-28
Outcome
Miss
Importance
2 / 5
Effect on score
-0.89

What counted as right

Status
resolved
Settled on
2017-09-28

The S&P 500 closes at least 4% below its 2017-08-28 close (about 2,444, i.e. at or below roughly 2,346) on at least one trading day on or before 2017-09-28. A decline of 2%-4% = PARTIAL; no decline of that size = FALSE.

Measured strictly against the written criterion. Reference: 2017-08-28 close 2,444.24 (FRED and Yahoo agree exactly). Required for TRUE: at least one close at or below ~2,346.47 (-4%) on or before 2017-09-28 — did not occur. Required for PARTIAL: a decline of 2%-4%, i.e. a close at or below 2,395.36 — also did not occur. Maximum peak-to-trough move available in the window: the minimum close was 2,446.30 (+0.08% versus the reference, i.e. the index did not post a single lower close), and the minimum intraday print was 2,428.20 (-0.66%). The index instead finished the window at 2,510.06, +2.69%. Deepest drawdown by either measure is under 1%, far short of even the PARTIAL band. resolved_on set to the criterion's deadline (2017-09-28) because the outcome only became final when the window closed without a qualifying decline. Evidence is unanimous across two independent sources; no conflicting readings. Per instructions I am not assigning an outcome value here, but the measurements above are stated so the quantify stage can apply the rubric directly.

Sources

  • 'The S&P 500 is set for about a 4 percent to 5 percent decline,' Tom Lee says

    X (Twitter) — @thesheetztweetz (CNBC) ·

    Source

  • History is pointing to a 5% drop in the market over the next month

    X (Twitter) — @fundstrat (Tom Lee) ·

    SourceArchive

  • Archived capture of twitter.com/fundstrat timeline

    Wayback Machine ·

    Source

Evidence of what happened

  • FRED series SP500 (Federal Reserve Bank of St. Louis, data from S&P Dow Jones Indices) gives the 2017-08-28 reference close as 2,444.24. The 4% trigger level is therefore 2,346.47 and the 2% level is 2,395.36. The LOWEST close anywhere in the 2017-08-29 to 2017-09-28 window was 2,446.30 on 2017-08-29 — ABOVE the reference close, meaning the index never closed below its 2017-08-28 level on any trading day in the window.

    Evidence

  • Yahoo Finance ^GSPC daily data independently reproduces every close in the window to the cent (2017-08-28 2,444.24; 2017-08-29 2,446.30; ... 2017-09-28 2,510.06) and adds intraday lows: the deepest intraday low of the entire window was 2,428.20 on 2017-08-29, just -0.66% below the reference close. Even on an intraday basis the decline never approached 2%, let alone 4%.

    Evidence

  • Rather than falling, the index rallied over the prediction window to a series of record closes: 2,500.23 on 2017-09-15 (first close above 2,500), 2,508.24 on 2017-09-20, and 2,510.06 on 2017-09-28 — a gain of +2.69% from the 2017-08-28 reference by the deadline.

    Evidence

  • Secondary corroboration of the 2,500 milestone date and level: Wikipedia's closing-milestones record lists the S&P 500's first close above 2,500 at 2,500.23 on 2017-09-15, matching FRED and Yahoo exactly.

    Evidence

Corrections

No public corrections.

Effect on the score

-0.89
How sure they sounded
85%
Outcome
Miss
Importance
2 / 5

This page shows the scored prediction as published; the site doesn't recompute or adjust the score.

Where this came from

The S&P 500 falls about 4-5% within roughly a month of late…
Methodology version
v1.0.0
Last computed
August 5, 2026
Made on
2017-08-28

How this profile was built

  1. Scored by independent judges

  2. Predictions gathered and sourced

  3. Independently verified against sources

Every public score traces back to sourced prediction records.

Avow is an independent project. The pundits scored here are not affiliated with Avow and do not endorse it. Names and likenesses are used for identification and commentary.

Evidence firstMethodologyPrivacy