# AI bubble burst by...?

DataSnack's model puts this at 24% seven days from now (10 Nov 2026); the market's number (Polymarket) is 23%. It has no graded forecasts yet; the first is graded on 12 Nov 2026.

## DataSnack's read

**AI's downturn risk has shifted from demand to who will fund the buildout**

A downturn would set terms for chip suppliers, lenders and the wider economy, so funding access matters beyond this market. The model's case rests on the chance decaying as the window shrinks, with a small yield tilt, at 30%.

## The numbers

- The market's number (Polymarket) now: 23% ([the market](https://polymarket.com/event/ai-bubble-burst-by))
- DataSnack's model, seven days from now: 24%
- The market closes: 1 Jul 2027
- How it resolves: This market will resolve to "Yes" if the AI industry experiences an industry downturn by the specified date, 11:59 PM ET. Otherwise, this market will resolve to "No". For the purposes of this market, the AI industry will be considered to have experienced an industry downturn once at least three of the following events have occurred within 90 days of this market's specified timeframe: - NVIDIA Corporation (NVDA) closing stock price is down 50% from its all-time high. - iShares PHLX Semiconductor ETF (SOXX) closing stock price is down 40% from its all-time high. - OpenAI, Inc. or Anthropic PBC…
- Record: It has no graded forecasts yet; the first is graded on 12 Nov 2026.
- Replay on the last 37 days: off by 1.8 points against 2.4 points for standing still; closer on 20 of 37 days
- Fitted on 42 days of history

## How the model works

The chance of a downturn by the June 2027 close decays as the window shrinks absent a new shock. The recent 10-year yield move adds a small data-fitted tilt on top of that decay.

- With no new break this week, the chance of a downturn by the close decays at 0 of the rate a constant hazard would imply. DataSnack's author expected 1; the data say 0, fitted on the same days the replay scores. A rough rule.
- Each one-point rise in the 10-year yield over three days moves the market's log-odds by -0.966 in log-odds over the following week. DataSnack's author expected -1.02 in log-odds; the data say -0.966 in log-odds ± 0.48, fitted on the same days the replay scores. Fits past data roughly.

## Sources

- AI bubble burst by...? · June 30, 2027: 23% on 10 Nov 2026
- 10-year Treasury yield (FRED): 4.86% on 6 Nov 2026

## The record

It has no graded forecasts yet; the first is graded on 12 Nov 2026.

Waiting for their day:

- Made 5 Nov 2026 for 12 Nov 2026: 42%
- Made 6 Nov 2026 for 13 Nov 2026: 29%
- Made 7 Nov 2026 for 14 Nov 2026: 28%
- Made 8 Nov 2026 for 15 Nov 2026: 28%
- Made 9 Nov 2026 for 16 Nov 2026: 27%
- Made 10 Nov 2026 for 17 Nov 2026: 24%

## Questions

**What is DataSnack's forecast?** DataSnack's model says 24% seven days from now.

**What is the market's number?** The market's number (Polymarket) is 23%.

**How does the model work?** It is an explicit equation fitted on data, with every link and source listed above. DataSnack's agents propose changes every night, and a replay against real data decides what is kept.

**How accurate has it been?** It has no graded forecasts yet; the first is graded on 12 Nov 2026. DataSnack's overall record is at https://datasnack.ai/accuracy.

**Can I download the model?** Readers who cover this market can export the model as an Excel workbook with working formulas from its page. A free example workbook: https://datasnack.ai/site/labs/datasnack-tariff-mortgage-lab.xlsx.

---

Canonical: https://datasnack.ai/m/ch:5e77e4799bec · Updated 10 Nov 2026 05:55 UTC · DataSnack forecasts are analysis, not investment advice.