Nov 142013

Then, on 29 July, Elan and Wyeth presented Phase 2 Bapineuzumab trial results at an international Alzheimer Conference in Chicago. A couple of months earlier, ahead of the Annual Shareholder meeting, Elan had organized its usual analyst dinner in Dublin, which I had attended. There was obviously a lot of interest in the upcoming presentation. The company would not comment, but the mood was distinctly positive. So much so that a month later, and a month before the Chicago conference, Elan and Wyeth released the top line results of the Phase 2 trial and announced the decision to go ahead with the third and final phase. After the announcement, as the stock price shot up above its post 2005 highs, I had decided to sell half of my position – a weight check discipline that I have come to respect more often than regret.

When the Phase 2 data were finally disclosed at the conference, they were basically in line with what the companies had announced a month earlier: not as good as one would have hoped for, but good enough to validate the decision to move on to Phase 3. And, surely, not as bad as to justify what happened the next day: a 42% price plunge, from 33 to 19 dollars.

Then – it never rains but it pours – just a couple of days later, as I was still trying to understand what had happened and what to do about it, Elan and Biogen disclosed two new cases of PML – the first after the 2006 readmission. Result: the stock dropped further, to below 10 dollars.

Holy cannoli. Now, there was no doubt that new PML cases would sooner or later appear. As by then 32,000 patients were on Tysabri worldwide, it was not a question of if but of when. But that it had to happen just after the Chicago presentation – not a choice: the new cases had just been confirmed and had to be promptly divulged – was a glaring case of the Luck of the Irish, or perhaps Murphy’s Law. And, of course, this double whammy had to happen right in the middle of the 2008 market cataclysm, to a company that was as far apart as any from financial turmoil and recessionary fears, and that – as I had kept smugly reminding myself – had appreciated by 50% since the beginning of the year. One of those things, I suppose.

With hindsight, it is easy to say that this was part of the risk I was taking. And of course there is a tendency to attribute positive outcomes to merit and negative ones to bad luck (Thinking, Fast and Slow, Chapter 17). But, in this instance, I disagree. My investment case in Elan was not based on the uncertain outcome of the Bapineuzumab trial, but on the solid reality of Tysabri efficacy. Clear evidence was there to conclude that there was very little risk of an escalation of PML cases, such as to discourage adoption or, worse, cause a new withdrawal. As current users approach 100,000, this proved to be right. It is impossible to demonstrate, but I think that, had the new PML cases arrived at any other time, they would not have provoked such a disorderly reaction.

And what about the reaction to the Phase 2 data? How could the market respond so violently to results that had essentially been preannounced a month earlier? This question brings us forward to the present, and to the insider trading case against Mathew Martoma of CR Intrinsic, an affiliate of SAC Capital. Martoma was at the Elan dinner in May 2008. I remembered his face when I saw him on the news, as a year ago he was indicted for obtaining inside information on the Phase 2 data, directly from the doctor who gave the Chicago presentation! Apparently, SAC Capital and CR Intrinsic were long more than 700 million dollars in Elan and Wyeth shares. On 30 June, with the price of Elan at 35 dollars, Martoma wrote an email saying he thought the stock would break 40 after the 29 July announcement. But after he was shown the Power Point presentation a few days before 29 July, the long positions were liquidated and 960 million in both stocks were shorted. Does this have anything to do with the ensuing price collapse? Who knows. If the accusation proves right, Martoma and SAC illegally benefited from it. But, as the shorting took place before the presentation, presumably they didn’t triggered it. Who or what did it, then?

According to one view, the Phase 2 data were such a “disaster” that the collapse was fully justified, and Elan’s and Wyeth’s decision to go ahead nonetheless with Phase 3 was fatally flawed. This view has been vindicated by the eventual failure of Phase 3, announced in the summer of 2012. But I think it is wrong to say that the data were negative – a view shared at the time not only by the two companies, which might have been ensnared in wishful thinking, but by a majority of the research community. Using our definitions, the evidence provided by the data was, however mildly, confirmative: LR>1. Judged more severely, perhaps one could say LR=1. But there is no way that LR<1.

As we know, however, evidence updates a prior probability. In this instance, the efficacy of the drug depended on the validity of the beta-amyloid hypothesis as the fundamental explanation of Alzheimer’s disease. If one assigns a high prior to the hypothesis, even slightly confirmative evidence can be enough to support it. But if the prior is low, support needs a higher degree of confirmation. Was the financial community sceptical about the hypothesis? They didn’t have a clue, of course. They were going along with the prior of the research community – that’s why Elan’s stock price went to 36 dollars after the June preannouncement. After the Chicago presentation, the research consensus was that the data were somewhat encouraging, or at most neutral. Only a minority – those who believe that the beta-amyloid hypothesis is false – saw the results as a confirmation of their doubts.

So why did the price collapse? I still don’t have a satisfactory explanation. Maybe it was a knee-jerk reaction, likely to be soon reabsorbed, if it hadn’t been immediately followed by the PML scare. But it is just a supposition. In any case, my valuation was based on Tysabri, not on Bapineuzumab. And with the price suddenly below 10, rebuilding a large position was again the right decision. It did take longer this time for the price to rebound: the financial crisis eventually caught up with Elan, and while its debt level had been well accepted until then, a heightened intolerance to leverage and a smaller capitalization forced the company to make several strategic changes. Over the following two years, the price fluctuated, falling at times even below 5 dollars. But as Tysabri continued its unabated progress, these all turned out to be buying opportunities: there was no doubt that the market would eventually catch up. And so it did, moving from 6 to 13 in the course of 2011.

2012 was going to be a big year for Elan. The outcome of the Phase 3 Bapineuzumab trial, expected in the summer, was going to shape the company’s future. In the meantime, Wyeth had been bought by Pfizer and, in order to redress its balance sheet, Elan had sold half of its Bapineuzumab interest to Johnson & Johnson. Nonetheless, a successful Phase 3 would have been a game changer for the company. Having bought below 10 in 2008 and again several times at lower prices in 2009 and 2010, I had recouped a good part of my double whammy losses.

So what was the probability of a positive Phase 3 result? The question hinged on the validity of the beta-amyloid hypothesis. What was the prior probability of the hypothesis? High, if you asked the Chief Scientific Officer of Elan, sitting next to me at the dinner in Dublin and passionately explaining why, of course, beta-amyloid caused Alzheimer. This is the Inside view, which he is now as passionately propounding as the CEO of Prothena. And he may well turn out to be right, eventually. By the way, Prothena’s share price has almost tripled since the company spun off from Elan at the end of 2012.

But the Outside view suggested more caution. The Bapineuzumab trial was not the first one based on the beta-amyloid hypothesis. Many had been conducted earlier, and all of them had failed. The negative evidence had been accumulating in recent years, thereby reducing the appropriate prior to assign to a Phase 3 success. The lower the prior, the higher is the accuracy of the evidence needed to support the truth of the hypothesis. Hence the excitement that had preceded Phase 2 had been replaced by growing scepticism. At 13 dollars, the market was giving a low probability to a successful Phase 3. So the payoff was highly asymmetrical: while a failure would have had a limited adverse effect on the stock price, a success would have yielded a substantial appreciation.

As it turned out, Phase 3 failed. The announcement came in two stages, on 23 July and 6 August, causing a cumulative 26% price drop, which took less than a year to be reabsorbed. Today the price is about 18 dollars, but on 29 July – five years to the day since that fateful Chicago presentation – the company announced it had been acquired by Perrigo.

Eleven years had gone by since my first 2 dollar purchase. What a ride.

Print Friendly, PDF & Email