Tuesday, 16 August 2016

Follow up Comments on Canadian Shareowner Ranking Screener, and Gilead at Top Spot

I spent a good chunk of time over the weekend thinking about possible shortcomings in my Shareowner screener (which I'm certain are plentiful). I noted that Gilead was ranked #1 in May and August according to my pseudo Greenblatt ranking system, and I made a general comment, as follows:
  • The entire ranked universe really is a relative point in time snapshot vs. the comparable stocks in the universe, and should be viewed as a starting point for additional research. For example, as at August 11th, 2016, Gilead had the highest ranking relative to every other stock in the Shareowner universe. The ranking tells us a set of facts, but does not tell us why this set of facts exists at this point in time. The job of the investor is to determine why Gilead is relatively cheap comparably. Sometimes, top ranked stocks are cheap for a reason.
With this said, I wanted to look further into the Gilead story as there seems to be no shortage of proponents for buying the stock based on apparent cheapness relative to other large cap companies in the S&P500, and especially based on relative cheapness vs. other large cap biotech or pharmaceutical companies.

While I personally don't like using P/E, it is a widely accepted conventional measure of current market based valuation on the fly. My problem with P/E is that it's a point in time estimate only, and in my humble opinion, investors can end up deluding themselves by solely relying on P/E without going beyond why a P/E is relatively low (or high).

I am by no means an expert on the biotech or pharmaceutical space, but my layperson's opinion on attempting to make relative comparisons within the space/s based only on P/E is to first understand the cyclical nature of the drug life cycle. There's a great article I read over the weekend illustrating the drug life cycle, link here, source, Awesome Capital.

If one were to use the above table of statistical probabilities through phase transitions and concurrent changes in valuation in order to broadly visualize how a sample company's P/E would look during each phase, I would suggest the following (I've added my own "later phase" italicized bullet points at the end):

  • Pre-clinical to Phase 1, little to no earnings, high R&D, extremely high to undefined P/E
  • Phase 1 to Phase 2, little to no earnings, high R&D, extremely high to undefined P/E
  • Phase 2 to Phase 3, little to no earnings, high R&D, extremely high to undefined P/E
  • Phase 3 to NDA/BLA, little to no earnings, lower R&D, extremely high to undefined P/E
  • NDA/BLA to Market Approval, initial earnings, lower R&D, high P/E
  • Introduction of product to market, generating incremental cash flows, substantial earnings, low R&D, falling P/E
  • Maturation through patent expiry, product continues to generate incremental cash flows, declining earnings, rising R&D, stabilizing P/E
  • Patent expiry, generic competition, falling cash flows, declining earnings, rising R&D, rising to high P/E
The article also outlines the connection between big pharma and biotech, as follows, which I found fascinating, a strategy which Gilead, interestingly enough, pursued to a tee (see excerpt below)

So here was Gilead back in 2011, purchasing a late phase HepC pipeline by way of an $11B deal, which, at the time, represented almost 1/3 of Gilead's market cap.

Returning to the apparent mystery of Gilead's relatively low P/E, my thoughts are as follows:
  • The market doesn't just arbitrarily assign a low P/E without reason, especially only two years removed from two blockbuster drugs being introduced to market. The fact that the P/E appears relatively low should be a signal that the market is telling us something about the expectations investors have pertaining to the company going forward
  • The patent protection on both Sovaldi and Harvoni, which accounted for close to 60% of F' 2015 revenue runs through 2029 and 2030 respectively, so there's a 15/16 year runway for incremental cash flows before patent expiry. A low P/E could be symptomatic of market concern over peak cash flows shortly after drug introduction due to pent up demand for initial treatment (or other possible reasons I'll outline below)
  • Peer group large cap biotech/pharmaceutical P/E's measured using Amgen, 17x, Biogen 18x, Abbvie 19x, Roche, 24x, and Novo Nordisk, 22x, average 20x. Applying a similar peer group multiple to Gilead, valuation would either have to rise to $327B from current (i.e., $16B net x 20x), or earnings would have to fall to $5.25B ($5B x 20x = $105B current mkt cap). 
  • The key point here is understanding the case for why earnings may more likely fall vs. a threefold rise in valuation as the overall justification for Gilead's below peer group P/E
Why doesn't the market believe the story?

I found the following articles, courtesy of the Chicago Tribune and Bloomberg, outlining the rationale supporting doubts over future growth.

First up, the Chicago Tribune, link here, summarizing background, current political backlash, and competitive threats to the Sovaldi and Harvoni story.

A few excerpts from the article follow:

In a nutshell (funny I just wrote that because I'm actually eating roasted peanuts as I'm writing this), there has been the equivalent of a political upheaval in response to Gilead's pricing strategy surrounding the introduction of both Sovaldi and later Harvoni. One only need peruse the 2015 10K in order to develop an understanding of how far reaching the consequences might be:

Next up, Bloomberg on competition from Merck, link here, and a few excerpts from this article, which I found fascinating, copyright Bloomberg:

The key takeaway I got from this Bloomberg article is that the case for Sovaldi and Harvoni peak sales seems plausible, even with Zepatier and Viekira each remaining at under $2B in sales through 2020, while Gilead's recently reported year to year variances in Q2 2016 Sovaldi and Harvoni sales seems to suggest the same.

Michael Price on Valuing Pharmaceuticals

How can I attempt to write a piece on attempting to intelligently address the mystery of Gilead's relatively low P/E without referencing Michael Price, who had the following to say about valuing pharmaceutical companies, courtesy of Greg Speicher.com, link here:

So here's my attempt at a Michael Price type valuation on Gilead. Assumptions as follows:

  • 5% decline in Harvoni and Sovaldi revenue between 2016 and patent expiry, no price cuts
  • All drugs facing patent expiry drop to 20% of previous year's revenues in the year after expiry
  • Assume 50% net margins
  • Perpetual cash flows subsequent to 2030 are discounted at 8% 
  • Discount cash flows by year at 12%, compare to current market cap
Results as follows:

  • Cash flows don't take future pipeline &/or drugs under development into account, they only consider current pipeline
  • Cash flows model Harvoni and Sovaldi out to 2030 at linearly declining rates (5% per year). Reality may prove much different, if Gilead bows to pressure and reduces prices with no accompanying increases in patient uptakes. 
  • The market is finite. According to the WHO there are approximately 3-5M US patients with Hepc. At close to 200K treatments sold per year (Harvoni + Sovaldi combined), and no competition from other drugs, courtesy of Merck &/or Abbvie, the US patient population would be close to treated by 3030 (or at least the %ge of the population who can afford the treatment).
  • Is a 50% net margin over time realistic? In years where R&D spend climbs, margins will be lower
  • On the subject of why I used a 12% discount rate, the answer is two-fold, 1) there is uncertainty over what form future sales growth will take impacted by price sustainability, future patient uptake, or competition, & 2) I believe that during R&D heavy years, net margins will be lower, therefore I upped the discount rate
Based on the above, current capitalization at $105M exceeds the sum of my estimated DCF's by around $13B, or just over 14%, and based on Michael Price's comments above, he seemed to only be interested in buying when his DCF's up to patent expiry exceeded current valuation (i.e., he had a margin of safety): in this case, he was buying at a discount and getting the pipeline for free. 

I don't believe this is the case with Gilead currently, and certainly, this is an interesting and counter-intuitive result, because it suggests that Gilead is currently overvalued with a P/E of less than 7x ttm earnings.

Update, Aug 16th, figured I'd add a monthly chart for perspective. The chartist in me asks whether this setup is symptomatic of a base being formed (the answer is no):

No comments:

Post a Comment