Skip to main content

Only Two Out of Ten Drugs? Really?

You sometimes see the claim that only twenty per cent of drugs that make it to the market actually recoup their development costs. Here’s a recent sighting, and there are more. This figure has never seemed quite right to me, though.

There certainly are drugs that don’t earn it back – of that there can be no doubt. But 8 out of 10? Keep in mind that we’re not talking about failure rates during development; these are marketed drugs that don’t turn a profit. (There’s little doubt about clinical failure rates, because both clinical trial registrations and FDA approvals are public knowledge). But what’s not public knowledge are the development costs for a given drug, making profit and loss calculations impossible on a case-by-case basis.

Here’s a post at Lyman BioPharma Consulting on this topic, and it’s well worth a look. Stewart Lyman makes several good points, including:

1. The biopharma industry spends more on R&D, as a percentage of revenues, than almost any other. I’ve talked about this as well, but nothing seems to stop the “Drug companies spend more on marketing than R&D!” folks. And don’t forget the companies that have no revenues (yet) who are spending like crazy on research, too.

2. At the same time, the drug industry has been highly profitable. Some of the analyses in this area suffer from survivorship bias and similar effects, but the fact remains that if your drug company is actually earning a profit, it’s probably a very good one as a per cent of revenues. (If you try to achieve those profit margins without going through Point One above, though, then you risk turning into the next Valeant!)

So how does that two-out-of-ten figure add up? The answer is that it almost certainly doesn’t. Lyman has tried to track this number back to its origins, and he believes that it goes back to a 1994 study (which has some significant problems itself). In other words, this is almost certainly a garbled myth. It’s hard to imagine how the industry could function with the preclinical failure rates we have (very hard to quantify, but high, when you consider how many projects you’ve worked on in your career versus how many actually were recommended for development), the clinical failure rates (documented, somewhere in the high 80% to over 90% range depending on time period and therapeutic area), and then add an 80% not-even-profitable-once-approved rate on top of that. No, this business is hard enough already without making it look even harder than it is. BIO and the others using this number should back off of it – would they really like to try to document it? Didn’t think so.

14 comments on “Only Two Out of Ten Drugs? Really?”

  1. Petros says:

    the Grabowski and Vernon 1994 paper can be found here–1301-view-464, this relies on data from the 70s and 80s when, of course, many more drugs were approved than has been the case since

  2. exGlaxoid says:

    There are several reasons why the conclusion of 2 out of 10 drugs were profitable. They only looked at US sales, which at the time where the main drivers, but foreign sales do add up over time. When these drugs went to market, the time before patents expired and generics or other competitors came in was much slower than now. In the 1970-80 period some, if not most, companies had some ethics, eg, Merck, BW, Lilly, Abbott, and others had very reasonable pricing guidelines, but still made a reasonable profit (11% ROI from the paper), which was considered reasonable before Wall St. got so greedy.

    Also, there were more compounds in each area, so even if you had a good drug, new ones were always coming out that were better, so sales on some drugs tapered off faster than expected, and that was not usually planned in to initial drugs prices. People used to buy their own medicines, not insurance, so prices had to be in line with what people could afford, so they could not price drugs at $1000/pill, or people would not buy them.

    The key back then was that each company had to only get one really good selling drug to make all of its profits, and the rest were able to just break even, and the company could still do well. They gave Zantac as an example, which made $3 in 1983 or 84, and was the bulk of the profits for Glaxo at the time. Now there are many drugs making that much per year, but the blockbusters were able to fund entire companies back then, so they could spend more on R & D than they might have otherwise done.

    The key take home message is that drug discovery, development and FDA approval has become slower, more expensive, and gotten less predictable than in 1960-1980, so the costs needed on each successful drug is now expected to pay for itself many times over.

  3. Anon says:

    All you have to do is consider how the distribution is skewed around the mean for both costs and sales value. Then it is hardly surprising that 80% of approved drugs don’t even make their own R&D costs back, while the other 20% pay for themselves and all the failures plus profit on top.

    Another easy-to-verify fact is that around 85% of all global pharma industry profits are made from US price differences.

  4. Anon says:

    PS. For the last point, just allocate total industry costs by Rx volume, then profit per country is based on sales less per volume costs. You will find that almost all profit is made in the US.

  5. Daniel Barkalow says:

    Does it even make sense to assign development costs to a particular drug? If you charge all of the development costs on every drug that doesn’t make it to market to each drug that’s at all similar in target or structure that does make it to market, you could probably make nearly every drug look unprofitable, especially if you multiply count the costs when multiple drugs’ development efforts arguably benefited. Even if you’re not trying to mislead, it’s hard to split up costs reasonably, and awfully tempting to assign them in proportion to how much the company makes on the drug, except for blockbusters where this would be clearly nuts.

  6. Simpl says:

    The R&D costs are not all associated to NMEs. A fair amount is spent later on new indications, better dosage forms and combinations (yes, traditional blockbusters, but also onco). Depending on the promise of the NME, such add-on projects get started during the launch programme to increase the marketing potential. If the data you are measuring cut off at 1st US launch, these costs may get lost, or even be added to the marketing budget.
    Another figure I’d like to see better documented is the registration cost (global, not US) – though I fear the worst! However, lumping it together with marketing costs just feeds the claim that Derek wishes would go away.

  7. M Levy says:

    I’m not at all surprised that greater than 50% of approved drugs do not mapy for themselves. Is it 80%, 70%, 62%? In some ways, exact number is irrelevant. And on the methods, it only can be done by looking back at drugs approved 15 to 20 years ago to get a true determination, not those approved within the last 5 to 10 years. And on top of the many other lost costs in R&D, companies still make the bulk of their profits on the sales of a few golden geese.

    The drugs that don’t get approved tend not to be thought about except by those close to the company itself. There are many, many of these…hopeful me-too’s, mismatched to the market, bad marketing projections, loss of opportunity due to time lost in R&D, favorite drug candidate children within a company that goes forward and makes it to approval, competition that was under appreciated, new modes of treatment, drug withdrawals, government fees, penalties, fines, while also taking into account the lost money compared to investing in a basket of bonds and stocks.

    Nonetheless, whatever the exact percentage of profitable drugs, is there another industry that would allow such a high fractional financial failure rate, while still keeping upper managements in place for 5, 10 years of more?

  8. Derek Freyberg says:

    Just slightly OT, but I don’t know how many people have seen yesterday’s “Bizarro” (my local paper runs it). No mention of T*ring or even Shkr*li, so I think you have to be a little in the know for it to make sense.

  9. SteveM says:

    Re: “BIO and the others using this number should back off of it – would they really like to try to document it? Didn’t think so.”

    Corporate Cronies depend on the Big Lie. Their PR mouthpieces tell it often enough, the sycophantic business MSM plays the conduit, the Crony Politicos lash on because that’s where the money is and then the Lie morphs into “common knowledge”.

    It’s sorta like the now normative understanding that Americans are too lazy and too stupid to do STEM work, so only hordes of immigrants who really are the “Best and the Brightest” can save U.S. R&D.

    Those bogus numbers BIO is using are etched in stone…

  10. anon says:

    Paper also says “Over the past decade, the industry has added nearly 111,000 new, high-paying jobs”

    So, we can stop complaining.

  11. Morten G says:

    But doesn’t this blog regularly say that price per NME is about one to two billion? What percentage of drugs deliver that kind of profit during their patent lifetime?

    1. simpl says:

      Hi Morten- very roughly, any blockbuster should exceed this.
      Assume a patent life of 12 years from first introduction, production and general costs of 25%, and sales near 100% max. after 6 years, often still climbing until shortly before patent expiry.
      Then you have the last 6 years at > 1 x10^9$ p.a. – 25%, = 4.5 billion, which pays for the R&D, launch, marketing, and patent expiry plan.
      A big chunk is the nominal interest, since the payback comes at the end, and the peak payout was a decade or so earlier. The current low interest rates are really helpful in holding this cost low at present, But I haven’t seen low interest costs picked out in annual reports, or dancing in the stock exchanges: I haven’t even seen it mentioned. Maybe it was one effect behind the lift in Pharma stock prices around 2010 – 14.

  12. Kaleberg says:

    The $1.1B out of $2.6B as opportunity costs strikes me as an accounting fiction. Basically, you put anything you want into an accounting soup. The trick is usually taking it out later. For example, newspapers went on a merger binge and carried the acquisition costs on their books as good will. Good will is a lot like opportunity costs, but different, kind of an accountant’s wrench as opposed to a screwdriver. When the newspaper business started losing ad revenue to the internet, that good will hit the newspapers hard. They were all crying a blue streak about losing money, but their business of reporting news, printing it and delivering it was making money just fine. It was all that good will that was killing them.

    I’m guessing that the drug companies are using what they call Hollywood accounting to minimize their apparent profits. For example, Return of the Jedi which grossed over $475M, cost $32M to make, but never turned a profit. In Hollywood they do this to avoid having to pay royalties to various parties like the directors, writers, or actors. The drug companies are doing something similar, except their goal is to be able to poor mouth whenever someone suggests doing something about high drug costs. Some how or another, the CEO still gets paid.

Comments are closed.