MCD Earnings Recap: Cohort Central Tendency Held, CEO Commentary Faded The Move

The Pre-Print Cohort Read

We published the live MCD cohort this morning: 47 prior MCD prints + adjacent QSR analogs, 5-day forward median +0.4%, IQR [-2.6%, +3.5%], and conformal 80% band [-3.6%, +4.1%]. The dominant feature was U.S. comp-store-sales surprise: comp beat → 5-day +1.4% (n=29, hit-rate 62%); comp miss → 5-day -1.6% (n=18, hit-rate 33%).

What Actually Printed

Q1 2026 EPS of $2.83 vs $2.77 consensus (+$0.06 beat). Revenue $6.517B vs $6.531B (slight miss). Global comparable sales +3.8%. The print hit the cohort's bullish bucket on the comp number — the dominant feature surfaced in our pre-print read.

Reaction: shares initially traded +1% pre-market on the comp beat, then gave back most of those gains during the morning session as CEO Chris Kempczinski commented that the consumer environment 'is certainly not improving, and it may be getting a little bit worse.' By midday, MCD was approximately flat to slightly negative.

What The Cohort Got Right And What It Missed

The cohort correctly anchored the magnitude — the 5-day median expectation of +0.4% was right inside the actual realized morning range. The IQR [-2.6%, +3.5%] captured the realized volatility cleanly.

What the cohort didn't capture: management commentary on the consumer environment is a feature that doesn't show up in the structured data the cohort uses. The cohort's feature_importance ranks comp surprise, traffic-vs-price decomposition, and digital/loyalty trajectory — but consumer-sentiment language in the prepared remarks is unstructured and only enters the cohort post-call when the chart shape stabilizes 24-48 hours later.

• Pre-print 5d cohort median: +0.4%; comp-beat bucket: +1.4%
• Conformal 80% band: [-3.6%, +4.1%]
• Initial reaction: +1% pre-market on comp beat (cohort: directionally correct)
• Mid-day reaction: flat-to-negative on consumer commentary (cohort: didn't capture commentary feature)

The Calibration Lesson

MCD's print is exactly the kind of central-tendency outcome the cohort is designed to anchor. The reaction is small in magnitude, mixed in direction, and the realized return is sitting inside the IQR. That's a successful calibration even though the day-of price action 'feels' uncertain.

The unstructured-commentary feature is a known gap in the pre-print cohort — and the right reaction is to retrieve the post-call cohort 24-48 hours later, when the chart shape reflects the market's processing of the call narrative. That's when the multi-day continuation read sharpens.