2026-04-30: $RMD predicted +0.78%, did -4.11%

# Yesterday's Pattern Picks: April 30 On April 30, the model made 20 directional calls across equities. Eleven picks moved in the predicted direction, yielding a 55% directional hit rate. The portfolio average return was 0.14%, with individual moves ranging from +3.93% to −4.11%. This represents a modest day with mixed signal quality. The day's most significant miss was the hero pick: RMD (ResMed). The model predicted a bullish move of +0.78% but the stock declined 4.11% instead—an absolute error of 4.90 percentage points. This was simultaneously the worst-performing call in the set. By contrast, the best performer was NVO (Novo Nordisk), which gained 3.93% against a predicted 0.28%, overshooting the forecast by 3.65 percentage points. This outcome reflects a calibration challenge in the methodology this week. The RMD call shows the model overestimated conviction in a bullish setup while underestimating downside risk. The directional accuracy of 55% is slightly better than a coin flip, but the magnitude errors—both the RMD miss and the NVO surprise—suggest the magnitude estimates lack precision. The gap between predicted and actual moves indicates the model may be underweighting intraday volatility or missing key event-driven catalysts. This is a recalibration signal, not a confirmation signal.