Pattern A — priced inMA is the cleanest specimen on the board: 8 of 8 quarters cleared, +4.3% average surprise, and an average 5-day move of +0.03%. That is statistical zero. KO (+0.44%), V (−0.63%) and AMZN (−0.36%, on a +26.3% average surprise) sit in the same dead band. The market has stopped being surprised by these names — the beat is the base case, so clearing it pays nothing.
Pattern B — negative drift on good numbersMSFT is the standout: a perfect 8/8 record, +6.2% average surprise, and an average 5-day move of −5.13% — the largest negative drift on the board. META pairs the largest clean beats in the group (+19.9% average surprise) with a −3.18% average drift. Both are perfect beaters that have historically bled after beating. Big beats into rich positioning is where drift has gone negative hardest.
Pattern C — drift-up followerAAPL stands alone: 8/8, a modest +4.5% average surprise, and a +4.54% average 5-day move — the market has historically kept repricing upward after its print rather than fading it.
The thesisRead the table by column and the correlation you expect is not there. Beat rate and post-print direction are approximately uncorrelated. MSFT, META, AAPL, MA and KO share an identical 8/8 record and their drift outcomes span +4.54% to −5.13%. "They all beat" tells you nothing about the move.
The Street vs the modelOn several of this week's names, the model's estimate sits apart from published consensus — above the Street on AAPL ($1.92 vs $1.88), MA ($4.92 vs $4.77) and AMZN ($1.94 vs $1.81), and at or just below on MSFT ($4.18 vs $4.21) and META ($7.05 vs $7.13). That spread between the published number and the number the model reads as the real bar is where the post-print reaction gets decided, and it is not visible on a headline "beat or miss" chyron.