This is a true value play. The trading bots got confused by the headline and tanked the price. This company is trading at its cash value. Humans realize they screwed up. The analyst kept their predictions unchanged and the COA loaded up on the stock. Here is an AI slop summary. Basically, the drug worked. They gave it to the control group.. it worked again. This threw off the data. CMPX reported results from its COMPANION-002 Phase 2/3 trial for tovecimig in biliary tract cancer. The good news The drug worked on disease control metrics: Progression-Free Survival (PFS): Patients on tovecimig + paclitaxel stayed progression-free 4.7 months vs 2.6 months for control. Response rate improved: Tumor response was 17.1% vs 5.3%. Those are statistically meaningful signals. The bad news (what the market cared about) The trial did not show a clean Overall Survival (OS) benefit, which is often the gold standard in oncology. Why? Because 54% of the control patients crossed over into the treatment arm during the trial. That contaminated the survival comparison.