CAMBRIDGE, Mass. ( TheStreet) -- I missed the call on Ariad Pharmaceuticals ( ARIA). While it's true that no one is right with all their stock picks, it's little comfort because I greatly dislike being wrong and losing money. The only positive that can come out of missing Iclusig's toxicity and getting Ariad wrong is learning from the mistake and trying not to repeat it in the future. What did I miss with Ariad? I fell prey to confirmation bias. In general, when you're long a stock, you're more likely to dismiss neutral and negative news and/or spin it positively. The opposite is true when you're shorting a stock. In other words, once you make a decision to buy or short a stock, you tend to see all additional information as confirming your position. Obviously, investors should try to minimize confirmation bias but one can never completely escape its effects. This is why I like engaging the bear thesis when I'm long. It forces you to rethink or re-evaluate your investment thesis. There were certainly bits of information about Iclusig's safety profile that in retrospect were more serious than I originally believed. The changes made to the protocol of the phase III front-line EPIC study -- excluding more patients with pre-existing cardiovascular conditions -- was probably the most important signal I overlooked. I interpreted the changes made to the EPIC study as Ariad trying to get a cleaner test of the "dirty drug" hypothesis. I also felt the changes were not out of line relative to changes made to studies involving similar CML drugs. Interpreting these changes to the most important Iclusig clinical trial as relatively benign was a mistake. More skepticism was clearly needed. The lower dose of Iclusig introduced into the "Cortes" study at MD Anderson could have confirmed the elevated safety risk, but again, I saw nothing wrong because competing CML drugs are also often dosed lower than their labels prescribe. The changes made to the "Cortes" study should have given me more pause. What really threw me, however, was the lack of adverse events reported in the FDA's adverse event database. I assumed if Iclusig's safety risk was really significant, it would have showed up in the FDA database, but it wasn't (yet.) Instead, Iclusig-related adverse events showed up in the long-term followup from patients enrolled in the pivotal PACE study used to get the drug approved. I was looking in the wrong place for clues to Iclusig's safety.