Why pharma needs to shift to an ‘innovation constellation’ investment strategy

In the last two decades large cap pharma has evolved away from relying on innovation from a vertically integrated development model to drive growth. Instead, it’s increasingly embraced inorganic growth as the primary tool for driving portfolio growth.

However, the current model may no longer meet future growth ambitions. Revenue potential and research and development (R&D) productivity are both declining, and deal price was escalating until recent market correction. Across large cap pharma, R&D productivity has decreased sixfold since 2013. Fewer products are crossing the blockbuster (BB) threshold, defined as earning over $1 billion in a year, and it’s becoming increasingly difficult to achieve incremental benefit with significant improvement in standard of care across many disease areas. Moreover, pricing pressure and the approaching patent cliff will both contribute to the declining revenue stream and ability to transact.

Deal price reached a record high in mid-2021 and average merger and acquisition (M&A) premium increased from 55% in 2010, to 120% in 2019 with a 9% compound annual growth rate. The current sector downturn has experienced a decrease in deal volume and value. A sustained turnaround could re-ignite the same trendline we observed through mid-2021, given sustained industrywide pressures.

Additionally, future deal-making will face challenges from external factors that require changes in what and how we transact going forward. External drivers include global political instability, such as relations between the U.S. and China, macroeconomic instability due to inflation and interest rates, and legislation such as the Inflation Reduction Act.

To diagnose possible pitfalls with current models, and to inform what the future innovation model should be, we evaluated how innovation was sourced and the resulting success seen across nine of the top 20 pharma companies between 2005 and 2021. We are also introducing the concept of an “innovation constellation” investment strategy that leans on earlier-stage focused investment models to drive capital efficiency.

For this initial analysis, we evaluated nine of the top 20 largest pharma companies by market cap: Johnson & Johnson, Roche, Pfizer, Merck, AstraZeneca, Novartis, Bristol-Myers Squibb, Amgen and GlaxoSmithKline.

Across the sample set analyzed, we saw two archetypes for companies that have relied on inorganic sources as revenue drivers in their current portfolios. Archetype one relied heavily on inorganic sources as a primary growth driver, with 70%-85% of 2020 revenue coming from externally sourced assets. The second archetype had more revenue contribution from internally sourced assets, with only 35%-55% of 2020 revenue stemming from assets sourced externally.

The amount of business development (BD) dollars spent ranged from $49 billion (Amgen) to $174 billion (Pfizer) in 2005-2021, with late-stage deals—deals with the lead asset in phase 3 or marketed at time of deal—accounting for 46%-80% of the total spend.

To inform how companies should think about sourcing innovation into portfolios in the future, we evaluated the historical performances of different investment strategies. We measured deal success using three sets of metrics across the time horizon where deal impact could be expected:

• Revenue return—The ultimate objective but a lagging indicator of success, it requires inference from older data.
• Pipeline productivity—This leading indicator of success allows for more recent data but requires inference to translate to revenue return.
• Shareholder return—Less time is required to see an impact from deals on shareholder perceptions or a correlation with revenue and pipeline success.

Our research found that, across the metrics we assessed, companies that sourced innovation earlier in their life cycle generally had comparable-to-stronger returns:

• Revenue return—Assets sourced earlier in development demonstrated potential for superior revenue return, even factoring in development cost and development or regulatory risk. On a revenue basis, we observed higher revenue upside and forecast attainment with earlier- stage assets.
• Shareholder return—Companies that sourced BB drugs at an earlier stage tended to see higher shareholder return.
• Pipeline productivity—Companies that sourced assets earlier in development and saw progression through the pipeline—as opposed to buying assets that were already revenue generating—were rewarded with positive shareholder perception. Companies were more successful when early-stage assets were sourced as the lead assets, as opposed to when they were included as part of larger late-stage deals.

Our analysis measured return on investment (ROI) most definitively by looking at dollars returned versus dollars spent. We calculated revenue return by dividing the total revenue from assets included in deals by the total dollars spent on transactions in a defined time period. We then compared the metric across early- versus late-stage assets.¹ Total spend included costs of all externally sourced assets, including those that did not launch, to account for higher risk in early-stage assets.

We have seen a two- to three-times decrease in revenue return for late-stage deals over time. For example, return on late-stage deals was between 26 cents and 30 cents for deals between 2011-2016, as opposed to between 60 cents and 84 cents for deals between 2005-2010.

On the contrary, even after accounting for failure, investment in early-stage assets, including preclinical, phase 1 and phase 2 have historically yielded superior revenue returns than late-stage assets, including phase 3 and marketed. For every dollar invested, early-stage assets generated 87 cents revenue in the first four years post launch, as compared to 67 cents revenue with late-stage assets before accounting for development costs. After including development costs² ROI was still higher for early-stage assets compared to late-stage assets, at 69 cents versus 64 cents.

Companies generally focused and spent most investment dollars on obtaining late-stage assets, often out of necessity to bridge anticipated revenue gaps. This frequently led to overpaid premiums and sometimes suboptimal choice of assets, limited by what was available at the time. However, when looking at a longer time horizon, investment on early-stage assets may be a more efficient use of capital, despite having higher risk and incremental development needs.

Part of the differential revenue return may be explained by the dramatic increase in cost for more de-risked assets. However, our research also showed that revenue potential for early-stage assets may be significantly higher than late-stage assets when given sufficient time.

We looked at revenue over time, post deal, for all the launched assets sourced as part of deals in 2005-2013. Although early-stage assets on average did not produce meaningful revenue for three to five years post deals—given the extra time required for development—the growth seen was almost exponential. On the contrary, average revenue quickly reached its plateau with late-stage assets. By year nine post deals, average revenue generated by an early-stage asset was approximately two to two-and-a-half times that with late-stage assets. Some of the most transformative assets came into the companies’ portfolios in early stage, such as Keytruda (preclinical), Opdivo (phase 1) and Darzalex (phase 2).

This suggests many transformative opportunities were usually captured at the early stages of the development life cycle. What remains as late-stage targets is more challenging to acquire at later stages. This might be especially truly in a therapeutic area (TA) like oncology, where deal activities were disproportionally active in earlier stage.

Compared to late-stage assets, early-stage assets outperformed consensus forecasts, which were forecasted when the assets were in phase 3. The median percentage of forecast attainment for early-stage assets were approximately 1.6 to 2.6 times higher than with phase 3 assets.

When early-stage assets entered phase 3, consensus forecasts still tended to underestimate revenue potential. This leaves potentially greater alpha for companies who seek to move and acquire rights earlier. This could be due to the novelty of assets that were sourced earlier stage, including first-in-class, where analyst expectations might not be as informed. Analysts face limitations when deciding when and what to act on, so there’s a high potential for earlier-stage assets to be overlooked given uncertainties in commercial potential.

In addition to looking at revenue return, our analysis explored other leading indicators that might suggest early success with an early stage leaning strategy.

We found that companies that were able to source BB drugs in their early stage had superior total shareholder return (TSR) over the decade compared to peers. These companies were able to capture hidden gems before they demonstrated proof of concept.

Most companies that were able to obtain BBs earlier in their development had annualized TSR of between 6.9%-8.4%. On the contrary, companies that obtained BBs later in development (or no BBs), had much lower annualized TSR of 2.7%-5.9% despite sometimes having a similar total number of BBs. Shareholder perception seemed to reward companies that invested early based on a well-planned, long-term story that fulfills the portfolio vision, rather than reactively transacting to fill revenue gaps with late-stage assets.

Overall, the success rate with externally sourced assets was slightly higher than benchmark probability of technical and regulatory success (PTRS), regardless of the phase the assets came in. This may be explained by rigor in BD target selection or dedicated effort in progressing external assets given high investment.

To understand the overall impact of external innovation on pipeline throughput, we calculated a composite asset progression score across asset phases from inorganic sources. This score accounts for the percentage of assets successfully launched from each phase, with heavier weighting towards earlier stage asset success as the probability is lower.

Our research found that the asset progression score was generally correlated with overall company success, as measured by annualized TSR. This suggests that shareholder perception rewards ability to source and progress high-potential assets early, as opposed to buying revenue from marketed assets when the upside might be capped.

Interestingly, early-stage assets were less likely to progress to launch if they came in as part of a late-stage deal as opposed to the lead asset in early-stage deals. For phase 2 assets, 23% of those sourced as the lead asset progressed to launch, as opposed to 17% when they were part of late-stage deals. Given the significant upsides observed with early-stage assets, companies can potentially capture some quick wins in improving capital efficiency by redirecting adequate resources to early-stage assets acquired as part of late-stage deals.

Large cap pharma has historically focused investments heavily in scouting de-risked, late-stage programs. Our research shows that it is not obvious that a late-stage dominant strategy yields stronger returns, and that early-stage assets may offer sustainable growth.

We saw assets sourced at early-stage offer tremendous revenue upside potential. In addition, they often outperformed analysts’ expectations, which is a proxy for market value, translating to potentially under-valuing. Ultimately, early-stage assets yielded better long-term revenue return even after accounting for failures and incremental development cost. Shareholder perception also seemed to reward companies with strategic foresight in investing high-potential opportunities early and progressing them through the pipeline, instead of relying heavily on acquiring marketed assets to fill immediate revenue gaps.

While it’s riskier to source external innovation earlier in their development life cycle, companies will need to move to earlier in the external innovation ecosystem to seize transformative opportunities. We believe a shift to a disciplined and intentional innovation constellation strategy, with right-scaled, early-stage activities, can help to improve capital efficiency and drive long-term success. This will include better screen and scout early-stage opportunities on the basis of risk, and rescaling organizations to embrace an early stage leaning BD strategy. We will fully explore these strategies in future articles.

1. Revenue includes the initial four years of revenue post launch. Given late-stage deals often included early-stage assets, we have decomposed the cost of such deals across assets at different stages of development, based on PTRS ratio and incremental cost of capital. Development cost was added to each asset based on the phase they came in and the highest phase achieved, with benchmark development cost by phase and TA. In order to allow sufficient time for early-stage assets to progress to launch, we have evaluated deals conducted during 2005-2013.

2. Development cost estimated for all externally sourced assets, based on the phase the asset was sourced and the highest phase reached. This includes, for example, if an asset was sourced at phase 1, the cost of phase 1 and beyond are included regardless of whether an asset was at beginning or end of phase 1. If the asset was abandoned at phase 3, the cost of whole phase 3 was included regardless of if trial was stopped at interim. Cost estimate is based on TA level benchmarks by phase from Evaluate Pharma (accessed July 2022).