Wednesday, September 10, 2014

Whitney Tilson says to short Exact Sciences ... Should you?

Whitney Tilson has released a paper on his "new big short idea". 2nd Update: Mr. Tilson's slide show is public again, so I'm removing the embedded images, refer to link in first sentence.

The first question, is it really new? I can't say for sure, but you be the judge.

At last year's value investing challenge the 3rd place winner Travis Cocke was featured on CNBC. Travis starts at 1:45 in the video and it's worth watching, it's pretty funny if you ask me.  Closing price of EXAS on August 30th, 2013 when this aired live, $11.57.  Whitney Tilson happened to be one of the judges of that "investment challenge".

Then, less than two hours after Seeking Alpha posted a piece that has turned out to be a catastrophe of incorrect assumptions by Gravity Research on April 30th 2014, Whitney posted this comment on SA:

The stock price closed at $12 that day.  On a side note the long and lustrous history of Gravity Research, an anonymous organization, spans less than a year and it is worth checking out their prior work.

In the current slideshow presented at this year's Value Investing Congress in NYC, Mr. Tilson presents 47 slides in a slide-deck before he gets to Exact. What does it have to do with Exact or his case against them?  Not sure.  Can't tell if it was part of a general presentation, or all meant to build "cred" for his Exact presentation.  But something did jump out at me.  One of the charts that is presented is for Raptor Pharmaceuticals (RPTP).  Which just happens to be another company that Gravity Research wrote a piece on.  Interesting.  Oh, Whitney Tilson also wrote an article on shorting Montage Technologies (MONT), after they had announced a buyout offer (?).  After Gravity Research wrote a piece on them as well.  Interesting.  None of these shorts seem to be working out very well for these fellows, if you consider price they went short, the price now, and their targets.

I don't think this is a new short idea at all.  I think this short position is old, stale, and the ideas in Mr. Tilson's paper are rehashed. With over 30% of the float of EXAS held short as of last report and over 97% of shares owned by institutions, very crowded and hurting.

Another thing that stuck out in the presser, was this blunder.  Drug? What Drug?  Geez, do people peer review these things?

Okay, on to the meat.

1) Mr. Tilson brings up the old laughable "ick factor" of fecal testing on Slide 53, but seems to go a step beyond what I have seen written elsewhere and makes the outlandish claim that FIT requires *less* handling of stool.  Amazing.

Let us review the differences.

-With iFOBT the patient must use a tool much like a stick to extract some stool from the toilet and wipe it on a card between 1 and 3 times, depending on a given test kit.  Studies have shown that compliance diminishes with more stool samples needing to be wiped on these cards (and the cards kept around for the next wipe before mailing)

-With Cologuard the device is placed over the toilet seat and you poop in the container, then pour a bottle of DNA buffer into the container and put the lid on.  There is no handling of the stool at all.

On the same slide, it is questioned if doctors and patients will really embrace the Cologuard platform. Survey says... Yes. On Slides 33 and 34 of Exact's presentation, it makes it clear that over 90% of physicians surveyed would be receptive to prescribing the test and 74% of FIT tests would be supplanted amongst surveyed physicians.  This survey was conducted in 2013 before FDA approval.

2) Mr. Tilson applies the kitchen sink method in Slide 54 to throw in every potential competitor there is, even including Epigenomic's product EpiProColon, which went before Advisory Committee at the same time as Cologuard and had a much worse recommendation profile compared to Cologuard and was subsequently sent back to do more work by the FDA.  I am not going to review all the blood tests and pillcam as that would be laborious, but at the end he throws in a curve ball with a seemingly amazing mRNA marker that blows Cologuard out of the water, abstract here.  What is left out, is that he is comparing the results of a clinically robust phase 3 trial (Deep C), with a tiny validation study.  And the real kicker that gets left out, is this , this mRNA marker is found to be upregulated in a wide range of solid tumors, including lung, breast, gastric, colon, and glioblastoma.  He's comparing this to a highly automated lab ready test.  Pretty funny, really.

3) They used the wrong FIT test!  Insure FIT is great! (Slide 52) Again, it is about comparing apples to oranges.  I would love to see how Insure FIT, a test that is more difficult to comply with than PolyMedCo's test, stacks up.  Maybe Insure FIT's maker could fund a robust trial, rather than this rather wimpy trial in which they stacked the deck with 161 symptomatic cancer patients out of 2512 participants?  By comparison the Deep C trial was made up of entirely average risk non-symptomatic patients, and only 67 cancers were found out of 9,989 patients with fully evaluable results.  An excellent non-biased comparison of screening methods available here.


As is usually the case, the devil is in the details.  In this case, many of the details left out of Mr. Tilson's presentation.

- One very important detail left out, is that parallel to FDA Approval, Cologuard also received a Medicare national coverage determination to be reimbursed every 3 years, and the FDA voluntarily improved the label from what was presented to the Advisory Committee, increasing the age cutoff to 85 and including language about pre-cancers.  The importance of this cannot be understated, as the company has worked with CMS and the FDA in this landmark parallel review process all along the way and is the programs first success story.  Cologuard is the first product to ever receive tandem FDA approval and CMS coverage.

- One Third of cancers develop from flat lesions known as Sessile Serrated Adenomas.  Cologuard's detection of these lesions is the same as for regular Adenomas.  FIT however is blind to them, with clinically insignificant rates of detection, because they do not bleed.  Colonoscopists also have difficulty finding these polyps, they are easy to miss.  It is a case where Cologuard could be highly complementary, even to those getting their once a decade scope, at the 5 year interval.

- Tilson questions commercial adoption uptake and cites the Aetna paper released prior to FDA approval, but fails to point out that the Mayo Clinic has already adopted it as a reimbursed screening test.  Some serious underestimating of the significance of this seems to be at hand in the short case.  Also, Kevin Conroy stated in the latest conference call that the economic models performed by a third party show Cologuard to be substantially efficient, and to beat both breast & cervical cancer screening in terms of quality adjusted life years, hands down no contest.  CMS is in receipt of this data.  I encourage you to listen for yourself.

The purpose of this article was not to valuate EXAS. If you are interested in what I consider to be well modeled and thoughtful reasoning on that subject, allow me to point you here

Full Disclosure: I am long shares of Exact Sciences.  I wrote this article myself and received no compensation.  Have a great week :)