Weekly category decision audit.

This page separates the model's category choice from the execution ticker. The category earns the overweight through final eligible category score; the representative ticker is then used for implementation. Final category score is no longer a simple three-ETF average. It is a proof-burden score that starts with the 3/2/1 weighted ETF basket, then tests leadership, volume-price sponsorship, persistence, timing, risk/reward, setup quality, and whether the active macro stance is favored, neutral, or a headwind. The only user-facing scorecard here is the four-week Monday-open-to-Monday-open holding window.

Top-2 performance by selected category

This table asks a stricter question: when a category is actually selected for an overweight sleeve, did that selected representative beat the average of the eight non-overweight category representatives over the full four-week holding window?

Top-2 categories vs bottom-8 category leaders

This is the clean category-selection test: each week compares the simple average return of the two overweighted category representatives against the simple average return of the other eight category representatives. If the top-2 average is not consistently better than the bottom-8 average, the category selection process needs work. The headline spread above is the average of these weekly spread rows, not a portfolio return and not an annualized figure.

Allocated asset weekly performance

Top-2 rows are the 30% winners in normal mode or 13% winners inside a 50% Defensive/crypto overlay. Asset return is measured from the Monday open after the report to the Monday open four weeks later when that forward return has been stored.