Records (VALID rows with 5y forward returns): 342

================================================================
 MATCHED-PAIR ANALYSIS
 Within fiscal year, control for fundamentals, isolate price
================================================================

  L1 LOOSE  — ROIC ±10pp, g_f ±10pp, any sector
    Pairs tested: 1368
    Gap-differential predicts: hits=685/1368 = 50.1% CI [47%, 53%] p=0.4892 
    MoS-differential predicts: hits=682/1368 = 49.9% CI [47%, 52%] p=0.5538 

  L2 MEDIUM — ROIC ±5pp, g_f ±5pp, any sector
    Pairs tested: 469
    Gap-differential predicts: hits=233/469 = 49.7% CI [45%, 54%] p=0.5733 
    MoS-differential predicts: hits=230/469 = 49.0% CI [45%, 54%] p=0.6778 

  L3 STRICT — ROIC ±3pp, g_f ±3pp, SAME sector
    Pairs tested: 24
    Gap-differential predicts: hits=10/24 = 41.7% CI [21%, 62%] p=0.8463 
    MoS-differential predicts: hits=14/24 = 58.3% CI [38%, 79%] p=0.2706 

================================================================
 + GAP-DIFFERENTIAL FLOOR (require |Gap_A − Gap_B| > 5pp)
 Tests whether the framework predicts when the price signal
 is loud, not just whenever fundamentals are matched.
================================================================

  L1 LOOSE  + |ΔGap| > 5pp
    Pairs tested: 826
    Gap-differential predicts: hits=402/826 = 48.7% CI [45%, 52%] p=0.7882 
    MoS-differential predicts: hits=398/826 = 48.2% CI [45%, 52%] p=0.8596 

  L2 MEDIUM + |ΔGap| > 5pp
    Pairs tested: 265
    Gap-differential predicts: hits=121/265 = 45.7% CI [40%, 52%] p=0.9299 
    MoS-differential predicts: hits=126/265 = 47.5% CI [42%, 54%] p=0.8051 

  L3 STRICT + |ΔGap| > 5pp
    Pairs tested: 9
    Gap-differential predicts: hits=2/9 = 22.2% CI [0%, 56%] p=0.9805 
    MoS-differential predicts: hits=5/9 = 55.6% CI [22%, 89%] p=0.5000 

  L1 LOOSE  + |ΔGap| > 10pp
    Pairs tested: 414
    Gap-differential predicts: hits=203/414 = 49.0% CI [44%, 54%] p=0.6708 
    MoS-differential predicts: hits=198/414 = 47.8% CI [43%, 53%] p=0.8248 

================================================================
 INTERPRETATION
================================================================
 - If Gap-differential accuracy > 50% with significance after
   matching, the Brina Gap adds price-discovery value beyond
   what fundamentals already predict.
 - If accuracy ≈ 50%, the Gap signal is noise once fundamentals
   are controlled for; the framework's univariate accuracy
   reflects fundamental differences, not the price signal.
 - Compare Gap accuracy vs MoS accuracy at each level — they
   should diverge if our earlier finding (MoS is at chance) is
   robust within matched pairs.
