Pipeline Success Rates Correlate With Biotech Valuations

The TD Cowen Insight

We investigated whether pipeline success rates have an impact on biotech industry valuations. We quantified Phase III success rates for 2005-2020, including a proprietary analysis from 2014-2020. We found that Phase III success rates correlated with biotech EV/R&D valuations but only weakly with NBI performance. Success rates appear to be a factor in biotech valuations, but less so in the bull/bear cycle.

Biotech Boom-and-Bust Cycles

Biotechnology stocks are infamous for boom-and-bust cycles. Prominent among biotech investors is the idea that clinical trial successes and failures can bring the industry into and out of favor. This would imply both that pipeline success rates fluctuate over time, and that the market reflects these changes by adjusting biotech valuations broadly, beyond the specific companies affected. We sought to investigate whether the historical data support this theory.

Is There a Correlation Between Pipeline Success Rates and Valuations?

We quantified aspects of R&D in biotech for the period of 2005-2020 including clinical success rates and R&D spend. We found that biotech valuations, specifically as measured by the ratio of enterprise value to R&D spend, increase during periods when clinical trial success rates are high. Moreover, we found that EV/R&D is depressed during periods when trial success rates are lower. We found little correlation between the NBI’s y/y performance and Phase III success rates during the period studied. Thus, it does seem that clinical success rates are a factor in biotech industry valuations, though perhaps not to the determinant of the bull and bear cycle.

Using our proprietary analysis for the period between 2014-2020 we found that probability of Phase III success correlated with EV/R&D valuations (R2=0.45). We found a very similar correlation in the period between 2005-2015 with the probability of Phase III success estimates derived from the Wong et al. academic paper (R2=0.46).

We found only a weak correlation between Phase III success rates and NBI y/y performance in both our dataset (correlation R2=0.18) and in the Wong dataset (correlation R2=0.02).

Methodology to Evaluate Historical Clinical Trial Success Rates

To evaluate historical clinical trial success rates, we examined data from two sources. For the period between 2015-2020, we performed our own new and proprietary analyses utilizing 1,240 Phase III events from the FactSet Drugs database and list of FDA approvals from CDER and CBER made available by the FDA. For the period between 2005-2014, we obtained data from an academic paper published in Biostatistics in 2019 by Wong, Siah, and Lo. We performed proprietary analyses to correlate Phase III success rates by year with industry valuation metrics and historical stock performance obtained from the Bloomberg database.

AI Assisted Clinical Trial Success Leads to Strong Valuation

Our analysis would suggest that if Phase III success rates improve, biotech valuations would increase. There is some hope among investors that an increased use of biomarkers and/or artificial intelligence could increase the success rates of clinical trial programs over the coming years. It would seem quite possible that better trial outcomes due to improvements in technology could potentially drive stronger valuations. Moreover, since there was a correlation between EV/R&D and probability of a Phase III product being approved, we can try to predict the likelihood that the Phase III programs that read out in 2021-23 will ultimately secure FDA approval based on the average EV/R&D ratio of those years.

Phase III to Approval Success Forecast

Using the linear regression equation, our data would forecast 2021-2023 Phase III to approval probability of success of 56%, 51%, and 47%, respectively. It will be interesting to evaluate the accuracy of these predictions once sufficient time has passed to allow all the Phase III programs of the past several years to fully progress through the regulatory approval process.