October 5–9, 2026

This week’s forecasts.
Against the Street.

Our call before the earnings call. Explore this week’s Opus 5.5 forecasts, compare with the Street, and open the models behind the numbers.

See all 11 forecasts

New forecasts every week

Pick a company. Open the forecast.

This week’s forecasts

Normalized EPS · USD/share

Made with Shortcut · Opus 5.5 · High effort

11 of 11 forecasts published.

October 5–9

10 / 10 companies scored

10wins
0winsWall Street

Reports released October 5–9. APLD and LEVI use direct numerical comparisons; see company-page basis notes.

0 still to score · Select a block to explore

24.0% lower

Shortcut’s median EPS error vs. the Street

Open October 5–9 forecasts and results ↗

Furthest from Street

Normalized diluted EPS, USD per share. SAR uses adjusted NII per share. Street estimates: snapshot as of October 4, 2026. A dash means a value is not yet available. Reported results use issuer-adjusted EPS, with sources on each company page. Scores compare the saved published estimates with reported adjusted EPS. APLD and LEVI use direct numerical comparisons, with EPS-definition limitations explained on their company pages. APLD and NRIX percentages show numerical forecast gaps; their EPS definitions remain unverified. NRIX is a conditional forecast. Model-specific assumptions are explained on each company page.

September 28–October 2

14 / 14 companies reported

9wins
5winsWall Street

Closer to actual normalized EPS. One block per company.

Open the full week’s results and models ↗

14.3% lower

Shortcut’s median EPS error vs. the Street

Open September 28–October 2 forecasts and results ↗

Weekly archive

October 5–9, 2026Current weekSeptember 28 – October 2, 2026Published

Our publishing record.

01

Before the report

We publish each prediction and its workbook before earnings. The original stays on the record.

02

After the report

We add the results and compare our EPS estimate with the Street’s, using comparable earnings figures.

03

The full picture

Wins and misses stay visible. Every weekly edition remains in the archive.

Research estimates, not investment advice.