Research Frame
Applies one intra-year window across the most recent 10 completed years.
Adjusts closed-market dates to valid trading sessions before comparing each year.
Treats the output as directional research context, not certainty or signal automation.
