Skip to Content

Business and Markets

Is Drug R&D Spending Going Up or Down?

I wanted to be sure to mention this piece by Frank David on drug industry R&D spending, not least because I’ve pounded away several times on that topic myself. He’s writing about a short paper in Nature Reviews Drug Discovery that he and Richa Dixit have published (you can contact him for a PDF if you don’t have access; his address is in the Forbes link above), which examines R&D spending trends from 2005-2015. Looking at the ten biggest R&D spending companies over that period, they find a Compound Annual Growth Rate (CAGR) for revenues of -0.01% (which is not a wonderfully encouraging number, let it be noted). The companies’ SG&A expenses, meanwhile, had a CAGR of -1.12%.

But R&D spending during the same period had a CAGR of +1.76%. The biggest companies are not cutting back on research – in fact, they’re continuing to spend more on it in the face of flat revenues. And because SG&A is going down, the proportion of money spent on research is going up even more (all these numbers are normalized for inflation). Another interesting thing was the reaction of companies whose revenues went down by 5% or more in a given year. The response to these shocks was to either continue to increase R&D or (at worst) to decrease it but still less than the amount that SG&A went down. The implication is that companies are protecting their R&D spending, whenever possible.

These are not the results you’d expect if you asked people outside the industry to predict them – in fact, you’d probably get some different estimates from people inside the industry, particularly scientists who have been through layoffs and re-organizations. I’d be interested to know how well this trend holds up if you move further down the list of companies, but for the larger ones, there you have it: R&D is considered important, and spending on it is continuing to grow, even after things that could easily keep that from happening.


22 comments on “Is Drug R&D Spending Going Up or Down?”

  1. Some idiot says:

    Hmmm… very interesting, and not at all what I expected… I have not read the article, but two questions spring rapidly to mind:

    (a) are mergers and acquisitions pigeonholed as R&D? If so, that would skew things a bit…

    (b) It would be very interesting to see the breakdown between spending in discovery, pre-clinical, phase 1, phase 2, phase 3, and post-launch label extension work. My gut feeling is that the observed increase in R&D is probably due to extra work done in the last few areas due to (a) a need to position new drugs better, and (b) probably longer and more complex clinical studies in order to satisfy the relevant authorities.


  2. Bill Ackerman says:

    But didn’t Valeant already prove that you don’t need to spend at all on R&D?

  3. drsnowboard says:

    Not entirely enthused by this sort of navel-gazing, but does in-licensed early assets come under R&D ? Because paying more for early assets / academic alliances / rights without the continuing infrastructure to actually translate that into clinical candidates doesn’t support the idea of this being healthy.

  4. Magrinho says:

    I’ll guess UP. 50/50 chance of being right 😉

    Pharma has rid itself the grinding expense of all those internal platforms and boring programs. R&D spending now more closely resembles stock picking, casino gambling and carnival games.

    Funds playing the long-short game must be making a killing right now.

  5. biotechtoreador says:

    Interesting. Comparing R&D spend from 10-Ks for 2011 and 2016 for PFE, MRK, LLY, and BMY one gets an increase in R&D spend from 26.4B to 28.2B, and increase of 6.5%, just shy of the 7% increase in CPI in same period. Clearly these 4 companies do not the whole mkt make, but interesting that bigger players not leading the charge.

    “R&D spending now more closely resembles stock picking, casino gambling and carnival games.”

    Of course, doing something that’s never been done always has an element of chance. Always will.

  6. Fluorinator says:

    Isn’t the second order question what counts as R and what as D and how much is spent on the respective buckets?

    If you for instance look at Lilly’s attempt to make their Alzheimer’s drug work, they probably burnt through a lot of money for all the clinical trials (which is D arguably) vs the actual process of lead ID/lead opt (what I’d consider R).

    1. Some idiot says:

      Precisely. And I would guess that this would skew the results a fair bit (see my comment above).

    2. NJBiologist says:

      Unless the skewness in definition changes over time, the use of a percent-of-original analysis will remove that skewness from consideration.

  7. Billy says:

    Having spent the last 20yrs exclusively on the “R” side of R&D, I still wonder why pharma lumps the two together. I suppose to a finance person, spending is spending. But I do wonder how much of this increased spending is a result of companies trying to gain approval for existing drugs in new indications. I’m sure the “R” portion of these organizations saw little to no increase in cash.

    1. LC says:

      I suspect R and D are lumped together due to R&D tax credits. It is easier for the accountants.

      1. Anon says:

        Maybe also because they are ultimately supposed to be on the same team, trying to achieve the same goal?

        Most of the problems I’ve seen in R&D are when R chucks crap over the fence to D just to make the numbers.

  8. anon the II says:

    So, I’m wondering: If Pfizer saw it’s spending increase by 5% annually on average and Wyeth, Warner Lambert, Park-Davis, Searle, Pharmacia/Upjohn, Monsanto, and a bunch of biotechs spending went to zero, would that be seen as an annual increase in spending?

  9. Belgian PhD student says:

    How does these numbers play out when you consider Eroom’s law? Is this increase in R&D spending increasing at the same rate?

  10. John Wayne says:

    Not all R&D spending is created equal. We’ve all seen really dumb crap being funded (huge up front deals with dubious small companies, massive clinical trials in highly competitive areas with a reasonable standard of care, and working on looser ‘me too’ cancer targets) while anything with a longer time horizon (novel research) being given the short straw. Don’t even get me started on the difference between the things you can wisely outsource, vs the things that the average C-suite manager thinks you can wisely outsource.

    1. Hap says:

      1) Short-term stuff is always easier to evaluate than long-term stuff – it’s easier to figure out the potential benefits (maybe) and easier to tell whether the research is progressing or not, and people will support the research correspondingly.

      2) Lots of management is only going to be there for a short time, so they need to choose things that will improve their companies’ positions in that time frame so that they get paid and can get a better job – anything else is not relevant. In theory, boards of directors and shareholders are supposed to look at for this, but they have many of the same incentives and can safely ignore those who have different ones. Lack of familiarity with the science and susceptibility to smoke and mirrors is a problem as well.

  11. David says:

    This seems almost elementary to me. Unless you’ve a product *guaranteed* to steadily bring in revenue at or above current levels, you’ve got to bring out new products, right? I mean Ford is always designing new cars, Boeing is always working on the next plane, next design innovation.

    Unless it’s temporary to get through a financial crisis, this is the last area you want to decimate, but even then, you don’t want to lose your human assets in the process, so you’d be very careful. Am I being perhaps sensible ie naiive here?

  12. Frank David says:

    Many thanks to Derek and the commenters above for their thoughts on our recent paper. There are many issues with these sorts of analyses that I’m all too aware of, but I just want to briefly address one particularly important point that was raised by a few folks:

    We’ve started looking into accounting regulations and/or conventions related to licensing payments, and so far we haven’t discern steadfast “rules” for where upfronts and milestones are recorded. (We’re slowly pulling our findings together into a blog post / article – stay tuned!) So, we share others’ concern that without knowing how individual pharma are recording L&A payments for R&D-stage assets on their books, it’s particularly hard to make too much of comparisons between companies.

    However, I and many others would argue that R&D dollars are fungible in terms of corporate finance – i.e., $1 spent on internal resources should be considered equivalent to $1 spent on in-licensing. If you buy that argument, then our R&D numbers should represent the floor of actual spending (i.e., not including L&A payments that weren’t attributed to the R&D line) – particularly if you look at the sum across all of the firms, which should be less influenced by the presence or absence of a few giant transactions.

    And even with its numerous methodological limitations, I think our paper’s broader point still holds. Many industry critics claim pharma companies are starving R&D despite reaping outsized revenues, based on analyses of R&D/revenues that don’t make much sense. (See Derek’s many prior posts on the issue.) Our data provide a different way of looking at pharma’s revenues and spending priorities, and at least to me, they suggest the story may be a little more nuanced than previously appreciated.

    Thanks, all, for your interest and your thoughtful comments.

  13. simpl says:

    The downside view also rings true. When overall revenue declines significantly, R&D spending falls over-proportionally.

  14. exGlaxoid says:

    My concern is that many products that are in-licensed are including premiums and other costs that were not part of the original companies research expenses, so that the ultimate amount being spent on research is going down. The amount spent on development may be going up, but the costs per drug are skyrocketing, so that means that ultimately fewer trials are being done, as the cost per trial goes up exponentially.

    The best thing we could do to control drug costs would be to go back to smaller phase 3 trials and then monitor more people in phase 4 trials. Many of the longer phase 3 trials have not caught the side effects that they are supposed to find, and in fact often find statistical noise than can be interpreted wildly and incorrectly. Phase 3 should be looking for real side effects and dangers, not the lower chance risks that people found in Avandia type drugs, which are hard to tell from noise.

    Also, if the FDA would leave more “questionable” drugs on the market longer, like Vioxx, we could better find improved versions of them, just as penicillin (which has large risks due to allergies alone), aspirin, Tagamet, zocor, and Rezulin all lead to better drugs (2nd, 3rd, and even 4th generation drugs) over time, but once Avandia and Vioxx were branded as “dangerous” all new work in those fields pretty much died along with the chance of finding better compounds in those areas.
    As an example, my mother was on Vioxx, and it worked great for her, allowing her to walk again. No other drugs did as well, for her, and she was not able to keep as active once she was off it, which has lead to her physical decline. I think the small heart risk was acceptable in her case, given no cardiovascular disease ever in her family, and her heart being great. But her knees were terrible. So I would have liked to see the drug stay on the market for limited cases. Maybe a better one would be out by now if so.

  15. saul Richmond says:

    The industry is slowly starting to break out R&D expenses, however imperfectly. Gsk discloses spend on early stage drugs (roughly discovery to PIIa) as well as late stage and line extensions and I believe Novartis does something similar. Also it’s possible to break out licensing spend -at least for some companies. Celegene for example summarises “real” collaboration expenses in the form of milestones. Down the road it may well be possible to get a better feel for “R” and “D” as opposed to “R&D”

  16. brian mckay says:

    We need new meds to advance mankind

  17. Sam Brians says:

    A product supplier’s accompanying literature states that “We care about the success of your business. We will offer you profitable products and services that will help your business to grow. A treatment that works well at a low operating cost is your number 1 priority and therefore it’s ours too.” This statement strengthens the common belief that pharma and cosmetic companies often pay more attention to cost-savings when doing research and development, rather than focusing on delivering high-quality products to patients or consumers. But, compromises with quality should not be made and even if such companies tend to produce low-cost treatments, they still should strictly follow the GMP guidelines

Comments are closed.