Biotech Values

[iwfal]

Limits of biotech/biopharma growth? This article made a big splash about a month ago, which puzzled me at the time since it provided unclear theses, some questionable assumptions, and poor extrapolation of data. Normally I would write a short post detailing the failings of the article, but since I think the topic is worthy of (clearer) discussion I waited until I had time available to do a more thorough job.


Article's thesis:

Here is the thesis that it initially propounds:

Like many industries, Pharma’s business model fundamentally depends on productive innovation to create value by delivering greater customer benefits. Further, sustainable growth and value creation depend on steady R&D productivity with a positive return on investment (ROI), in order to drive future revenues that can be re-invested back into R&D. In recent years, however, it has become clear that Pharma has a serious problem with declining R&D productivity.

But in the conclusion, it partially undercuts its own thesis:

Just as the Pharma industry evolved from the chemicals industry, and the Biopharma industry has evolved from the Pharma industry, the Pharma and Biopharma industries together will evolve into something quite different, most likely continuing the historic trend of increasing complexity towards more complex biological solutions to pressing healthcare problems, such as cell & gene therapy, tissue engineering and regenerative medicine:

The confusion here is that the article does not clarify terms like ‘pharma’, or ‘biopharma’ or ... either in its theses, or later in the data analyses - instead focusing on putting out scary numbers. (It is always much easier to get a particular result if definitions are fluid.) Does current ‘pharma’ include ‘biopharma’? Does ‘pharma’ include small biotech? Which types? Certainly in my first reading of the thesis, and the first pass through the analysis, I assumed the article was asserting the combo of BP and all biotech was stagnating. It was only the final few paragraphs that seemed to be asserting something different – but what?

Data Analyses:

The key evidence it cites for the issues it raises are given in the 1st and 2nd figures, which indicate that the R&D Internal Rate of Return for “Pharmas” is falling off a cliff (the article (incorrectly) shows/implies a decline from close to 25% IRR in 1995 to 3% in 2017 and in a very linear fashion). Further in the same graphs, as support for this it cites other sources, such as Deloitte, which gives a ~3% Internal Rate of Return (IRR) in recent years. Finally, it should be noted that the 1st and 2nd figures are based upon data from EvaluatePharma (and the article doesn’t define what is in the dataset – another instance of the definition problem).

This seems a worthwhile discussion even if the article itself is seriously flawed:
a) Is the IRR really falling off a cliff?
b) Is it universal? Or does it only apply to some companies and/or technologies?

As I noted previously, the first, and most obvious, thing to do is to unpack the definitions. Although I couldn’t find a definition of what went into the EvaluatePharma dataset, there were some things that provide clarity:

a) The Deloitte data used in the article turns out be exclusively Big Cap Pharma (see Deloitte data ) and does not include “Medium Cap” Pharma. It turns out this is important since the “Medium Cap” analysis by Deloitte does not show the same (very poor) IRR. I.e. the Deloitte data that is included in the article is the piece that supports the thesis at the start of the article – and excludes uncooperative data.

b) Whatever the EvaluatePharma dataset includes, it is clearly a convenient dataset to show a peaking of revenues and all the IRR problems that might indicate. But there are plenty of other pharma revenue lists that do not show such peaking. (My guess is that one of the key parameters is whether the list is truly global, and another is whether small/medium cap is included – since global datasets and smaller cap both show continued growth in multiple datasets.)

But even if we take the EvaluatePharma data as a given, it is fairly clear that the extrapolation of the IRR (as derived from that EvaluatePharma data) out to current years (to align with the cherry picked Deloitte data) is incorrect. In particular, in order for the IRR to continue down the linear slope to 3% in 2017, the growth of EBIT would have to be substantially less than R&D growth in recent years and this has not been true in the EvaluatePharma data shown (as best I can tell from the crude graphs the growth rates from 2007 are equivalent). The linear decline of IRR that the article shows from 1995 through about 2004, that maps nicely to the Deloitte data, is largely due to an odd spike in the EBIT from 2004 to 2005 – and the subsequent data does not support a linear decline of IRR unless you postulate that the EvaluatePharma dataset is about to show a huge, sustained, drop in EBIT for 2017 and later (see my comments in the Caveats/Comments section of this post).

All told, I think the real story here is that, yeah, Big Pharma may have a problem (see caveats at bottom). But the rest of the ecosystem, including mid-caps, is still in fairly good shape. If anything this is an argument that Big Pharma has a really strong need to do acquisitions of smaller companies.

Caveats and Comments:

1) While the Deloitte data has more clarity than the EvaluatePharma data provided in the article, it is still fairly opaque – in particular how do they forecast expected revenues for 2017 R&D, … ? I’d suggest that inevitably any such extrapolation over such decadal timeframes is extremely problematic.

2) One of the closing arguments of the article is an analogy to oil reserves – e.g. showing we’ve already attacked the easy targets. Sure, there is truth to this – but only some, because very very few diseases are actually cured. Yeah, it isn’t very economically productive to chase CVD treatment now because the trials have become wildly too expensive and the societal cost too great (way, way too high an NTT). But there are still a *lot* of people dying of CVD – and all that is needed to access this is a better metric to predict who is at risk despite existing SOC (a pet peeve of mine has *long* been that companies will happily spend billions on huge trials even if spending $100M on a long term study to find good biomarkers would save them most of those trial costs).

3) The extrapolated curve of EvaluatePharma revenues etc (a pretty bell curve) at the end of the article takes as an assumption the thing that this article is supposed to be proving (a declining IRR). To date the EvaluatePharma does not indicate a decline in revenues (even assuming that EvaluatePharma is the dataset we want to be looking at – vs (for instance) the Deloitte mid-caps).

4) Interesting note – if the article had used the same methodology to show the IRR for the EvaluatePharma data back to 1990 it almost certainly would NOT have been on that nice, linear curve since a large cause of that nice linear decline is ‘recovery’ from the anomalous spike in EBIT growth from 2004 to 2005. I.e. all the article’s IRR data analyses always include this year –probably for this reason.