Veto edge = mean alpha of KILLed candidates minus mean alpha of PASSed ones. Negative is good — it means the veto blocked trades that went on to do worse than the ones it allowed. Measured against the Nifty Smallcap 250 rather than raw return, because the veto does not fire evenly across market conditions and a raw comparison would partly measure which weeks it happened to fire in. Needs outcomes filled in by the post-mortem, so it stays blank until decisions are ~21 days old.