The $2.7B Drug Problem: Pharma’s R&D Productivity Paradox

Pharma has never had more science to work with. AI is entering drug discovery, biologics have expanded the therapeutic toolkit, and the global pipeline still contains nearly 23,000 active drugs. Yet turning that science into an approved medicine is becoming extraordinarily expensive: according to Deloitte, the average cost of taking an asset from discovery to launch reached $2.67 billion in 2025, up from $2.23 billion in 2024.

The increase is particularly striking because higher spending has not translated into a comparable long-term improvement in productivity. McKinsey estimates that aggregate pharmaceutical R&D productivity has remained broadly flat since 2012. At the same time, Deloitte estimates that projected R&D returns across the late-stage pipelines of 20 leading biopharma companies increased from 5.9% in 2024 to 7.0% in 2025, marking a third consecutive annual improvement.

That creates an important question for investors: has Pharma finally started solving its productivity problem, or are a handful of exceptionally valuable drugs making the industry's underlying economics look healthier than they really are?

More Money for Every Shot on Goal

The economics of drug development are unforgiving because companies are not simply paying for the medicine that eventually reaches patients. They are also absorbing the cost of every molecule that fails along the way. Deloitte's cost-per-asset methodology incorporates both R&D expenditure and portfolio failure, providing a more complete picture of the capital required to produce a successful asset.

On that basis, average development cost increased by almost 20% in a single year, from $2.23 billion to $2.67 billion. Seventeen of the 20 companies in Deloitte's cohort increased R&D spending, even as their combined late-stage pipeline contracted by 4.6%. The implication is straightforward: more capital is being deployed against fewer late-stage opportunities, raising the economic hurdle each successful drug ultimately needs to clear.

As development becomes more expensive, every successful asset must generate more commercial value simply to compensate for the capital consumed elsewhere in the portfolio. A pipeline can therefore remain scientifically productive while becoming less attractive economically — particularly when the commercial value of that pipeline is increasingly concentrated in a small number of assets.

The Blockbuster Effect

That concentration is becoming difficult to ignore. Deloitte estimates that blockbuster drugs accounted for 70% of projected late-stage pipeline revenue in 2025, up sharply from 54% in 2024. Between 2020 and 2024, that share had fluctuated between 47% and 55%; the jump to 70% represents a significant break from the recent range.

Part of the explanation is the extraordinary commercial potential of GLP-1 and GLP-1/GIP assets. These programs now represent approximately 38% of projected commercial inflows from the late-stage pipeline analyzed by Deloitte. Remove them, and expected R&D IRR falls from 7.0% to just 2.9%, while average forecast peak sales per asset decline to $353 million.

The therapeutic mix has changed just as dramatically. For the first time in Deloitte's 16 years of analysis, obesity has overtaken oncology as the largest contributor to expected pipeline sales, accounting for roughly 25% of forecast sales, compared with just 1% in 2022. Pharma's improvement in projected R&D returns is therefore real in financial terms, but it is far from evenly distributed. A relatively small group of mega-blockbusters can materially improve industry-level returns even while the economics of the broader R&D portfolio remain challenging.

Innovation Is Still Reaching the Market

The industry's productivity problem should not be confused with a lack of scientific output. IQVIA recorded 79 novel active substances (NAS) launched globally in 2025, bringing the five-year total to 388. Based on the late-stage pipeline and historical success rates, IQVIA expects the industry to maintain roughly 70–80 novel active substance launches per year over the next five years.

That distinction matters. Pharma continues to generate new medicines at a substantial rate, even as the capital and time required to produce them remain under pressure. IQVIA reports that the recent improvement in clinical-program productivity was not sustained in 2025: trial durations increased and the interval between successive trials expanded by approximately three months, while end-to-end clinical development returned to roughly 10 years, among the longest timelines observed during the past decade.

Time ultimately becomes another form of R&D cost. Every additional month before approval delays potential revenue, consumes additional resources and reduces the effective commercial window before patent expiry. Improving productivity therefore does not necessarily require discovering more drugs; shortening the path from promising science to commercial launch can itself create significant economic value.

Innovation Is Moving Beyond Big Pharma

There is another structural shift occurring underneath the headline pipeline numbers. Pharmaprojects data show that the share of the global pipeline controlled by the industry's largest companies has fallen substantially over the past 15 years, while companies with only one or two drugs now account for a growing share of development activity.

The top 10 companies represented roughly 13% of the global pipeline in 2011, but only around 6% in 2026. By contrast, companies with just one or two pipeline drugs increased their share from roughly 15% to more than 22% over the same period.

For large Pharma, that fragmentation changes the productivity equation. Innovation increasingly sits outside the walls of the largest R&D organizations, making licensing, partnerships and acquisitions more important components of pipeline construction. The competitive advantage is therefore shifting from simply maintaining the largest internal research engine toward identifying external innovation early, acquiring it at attractive economics and allocating development capital toward the programs with the highest probability-adjusted commercial potential.

That becomes particularly important when the cost of carrying mediocre assets has never been higher. Deloitte finds that approximately 53% of late-stage asset-indications in its 2025 analysis have forecast peak sales below $250 million. With development costs approaching $2.7 billion on a risk-adjusted basis, companies have a growing incentive to terminate weak programs earlier and concentrate resources behind differentiated assets.

Can AI Bend the Curve?

AI is the industry's most obvious attempt to change this equation, but expectations should remain measured. The opportunity extends across target identification, molecule design, trial planning, patient recruitment, portfolio decision-making and regulatory processes. McKinsey argues that capturing those gains requires more than adding isolated AI tools; Pharma companies need to redesign technology stacks, workflows and decision processes around them.

There are early indications that this could eventually matter clinically. IQVIA found a 75% Phase I success rate among a vetted cohort of AI-enabled programs from emerging biopharma companies over the most recent three-year window, above comparable non-AI-enabled programs. Phase II performance remained broadly comparable, while there are not yet enough Phase III transitions to draw meaningful conclusions. IQVIA also cautions that the validated AI cohort remains relatively small, making it too early to conclude that AI has solved pharmaceutical R&D productivity at scale.

The more immediate investment case may be less dramatic but more economically relevant. AI does not necessarily need to discover the next mega-blockbuster to generate returns; technology that helps companies fail faster, select better assets and remove months from development timelines could materially improve R&D economics across an entire portfolio.

Bottom Line

Pharma's R&D problem is not a shortage of spending, pipeline assets or scientific breakthroughs. It is the efficiency with which those inputs are converted into commercially valuable medicines. Development costs have reached approximately $2.67 billion per successful asset on Deloitte's risk-adjusted measure, clinical development can consume roughly a decade, and an increasingly large share of expected pipeline revenue is concentrated in a relatively small number of blockbusters.

At the same time, the source of innovation is becoming more fragmented. Smaller developers account for a growing share of the global pipeline, making external innovation and capital allocation increasingly important alongside internal discovery. For investors, R&D productivity should therefore be viewed not simply as a scientific metric, but as a measure of how effectively management converts capital, time and pipeline optionality into economic value.

The companies best positioned for the next cycle may not be those with the largest pipelines or the biggest R&D budgets. The differentiator may instead be the ability to select differentiated assets earlier, terminate weak programs faster, source external innovation intelligently and compress development timelines.

In an industry where one successful medicine can still generate tens of billions of dollars, innovation remains extraordinarily valuable. The expensive part is figuring out which innovation deserves the next $2.7 billion.

Sources

1. Deloitte — Navigating the GLP-1 Boom: Measuring the Return from Pharmaceutical Innovation, 2026 — https://www.deloitte.com/global/en/industries/life-sciences-health-care/perspectives/navigating-the-glp-boom.html

2. IQVIA Institute — Global R&D Trends 2026: Advancing Innovation in a Changing Landscape, 2026 — https://www.iqvia.com/insights/the-iqvia-institute/reports-and-publications/reports/global-r-and-d-trends-2026

3. Citeline / Pharmaprojects — Pharma R&D Annual Review 2026 — https://www.citeline.com/en/resources/pharma-rd-annual-review-2026

4. McKinsey & Company — Strengthening the R&D Operating Model for Pharmaceutical Companies — https://www.mckinsey.com/industries/life-sciences/our-insights/strengthening-the-r-and-d-operating-model-for-pharmaceutical-companies