AI Scorecard: 50% Accuracy for the Week of July 14th

By Yogurt · 2026-07-18 · AI Scorecard

Our weekly review of Yogurt's AI-driven earnings predictions. This week the engine verified 6 reports and correctly called 3 outcomes — BALY miss, FITB beat, and RBCAA beat — for a 50% accuracy rate. Full breakdown inside.

The Yogurt AI engine reviewed six earnings reports for the week of July 14–18, 2026, correctly calling three outcomes for a 50% accuracy rate. The model nailed the banking sector — FITB and RBCAA both beat as predicted — but three regional-name surprises kept the score at the midpoint. Here's the full breakdown.

This Week's Results (July 14–18, 2026)

Here is the full breakdown of our predictions versus the actual reported outcomes:

Key Takeaways

• BALY (Bally's Corporation): The AI correctly identified weakness — Bally's reported a miss as ongoing integration costs and softer gaming revenues weighed on results.
• FITB (Fifth Third Bancorp): A clean beat. Fifth Third's Q2 net interest margin held up better than feared, and fee income surprised to the upside — confirming the bullish case for well-run regional banks.
• RBCAA (Republic Bancorp): Beat expectations with solid loan growth and disciplined expense management. Another accurate banking call from the model.
• ALV (Allianz SE) — MISS: The AI predicted an outright beat, but Allianz came in at exactly consensus — a "meet" rather than a beat. Solid underlying results but no upside surprise. The model over-estimated momentum in the European insurance segment.
• RF (Regions Financial) — MISS: The AI called a miss but Regions Financial delivered a beat, boosted by stronger-than-expected loan demand in the Southeast U.S. and better credit quality than consensus models anticipated.
• SPFI (South Plains Financial) — MISS: Another near-miss — the AI predicted a miss but the company reported in-line with estimates ("meet"). Small community banks are notoriously hard to model due to thin analyst coverage and volatile local loan dynamics.

What This Tells Us

This was a regional banking-heavy week, and the model split the difference. FITB and RBCAA were clean calls — the AI's advantage in large-cap banking continues. The misses (RF and SPFI) highlight the persistent challenge of modeling thinly-covered regional names where a single large loan or local economic surprise can flip the outcome. ALV is the most instructive: a "meet" technically isn't a beat, and the model's tendency to lean bullish on large European financials is a calibration note going forward.

At 50%, this week is below the model's long-run average — a reminder that not every week is a 67%+ performance. The Schwartz reveals all, but sometimes the market reveals it first.

Cumulative all-time accuracy: 63% (61 correct out of 97 verified predictions)

Track our live AI predictions inside the EarningsShot app. New predictions are posted every week before earnings reports drop.