2026-05-27: $DNLI predicted -0.47%, did +3.66%

On May 27, the pattern picks went 7 out of 20 directionally correct, a 35% hit rate. The portfolio averaged a 0.2% gain across the day. The best single move was DNLI, which gained 3.66%, while the worst was PG, which fell 1.07%. The hero pick was DNLI, flagged as bearish with a predicted decline of 0.47%. Instead, the stock rose 3.66%—a miss of 4.14 percentage points in magnitude. While the direction call was wrong, DNLI's actual move was the largest win in the set, which creates a somewhat inverted outcome: the methodology identified the stock that would move most, but called the direction backward. This kind of miss tells us something about the current state of the pattern-matching approach. The model is occasionally identifying volatile or significant-move candidates but is not yet reliably getting the directional sign right, even when magnitude signals are present. With a 35% directional accuracy over 20 picks, the method is performing below a coin flip, suggesting the bearish and bullish signals are not yet well-calibrated to actual next-day moves. No extrapolation should be drawn from a single day, but the pattern warrants watching how direction accuracy and magnitude tracking diverge or converge over the next several sessions.