Why rDNA.ai?

For more than four decades the biopharma industry has grown, matured and, over certain periods, survived via reliance on alliances. As an example, when Covid hit in 2020, one might have predicted that the robust biotech IPO window, then in its eighth year, might have gone into lockdown along with the world’s economy. Instead, 90 companies went public in the US in 2020, raising $18 billion in aggregate, or 3x as much as was raised in 2019 by the 52 biotechs in the IPO Class of 2019.

Why? It turns out that alliances, specifically “mega-alliances” ($100M+ in aggregate payments to the licensor) were the strongest correlate with successful biopharma IPOs. An analysis done by my former company, BioSci, showed that biotechs that went public from 2013-20 had $85 billion in potential alliance payments pre-IPO, plus an additional $115 billion from alliances commenced post-IPO – 4x the $50 billion in aggregate IPO proceeds raised over the eight-year period. BioSci also showed that one-third of all $100M+ alliances involved licensors who were members of the IPO Class of 2013-2020. (See BioSci analysis)

Since 2020, the biotech IPO window closed in mid-2022 after an impressive 122 companies raised an additional $18 billion in 2021. Yet biopharma alliances have continued to roll along, with $325 billion in potential alliance payments for licenses commenced over the 2021-23 period.

So, all is well for the biopharma industry, powered along by reliance on alliances … but all is not quite the same. Of the $325 billion in potential payments for recently announced deals, only $100 billion is covered by biotech alliances that have been SEC-filed. The remaining $225 billion (69%) are for deals disclosed via press release only.

After Four Decades, the Rules Suddenly Changed

As Covid took hold in March 2020, the SEC rolled out the first of three sweeping changes. Under the guise of work-from-home restrictions, the SEC canceled all pending Freedom of Information Act (FOIA) requests, including several hundred from my then firm, BioSci.

Secondly, via a memo sent to the US Archivist in February 2020, the SEC re-defined the retention period for documents subject to FOIA requests from 10 years post-expiry of confidentiality to three years from the date of filing – thereby permitting the immediate destruction of at least 5,000 unredacted biopharma contracts in the SEC’s possession.

And thirdly, via notices posted in the Federal Register in April 2019 and January 2021, the SEC permitted filers to redact confidential information from filed exhibits without submitting the unredacted exhibit. Moreover, because of the US Supreme Court’s decision in the Argus Leader case in June 2019 that “confidential” information need not be disclosed to the government (including the SEC) so long the submitter regarded it as secret, many public biotechs ceased to file even redacted versions of their alliances with the SEC.

As a result of these changes, the biopharma industry is far more reliant on the self-reporting of alliance terms than it has been over the past four decades. In BiosciDB, for example, there are fewer than 300 SEC-filed contracts for deals commenced over the 2021-23 period, versus more than 1,000 during the previous three-year period from 2018-20. Two-thirds of the most recent biopharma alliances were SEC-filed for deals signed in 2021, with a precipitous drop-off over the past two years.

What Artificial Intelligence Can Offer

For approximately 1,500 companies involved in the biopharma industry who file financial information regularly with the SEC, generative AI has the potential to fill the gap created by the absence of reliable data on alliance contracts. US-based public companies have filed interactive notes to financial statements, including descriptions of material alliances, on a quarterly basis for more than a decade. Non-US public companies that are cross listed on US exchanges have also filed with the SEC at least annually.

rDNA.ai has been formed to harness the speed and capacity of AI to analyze and update alliance-related disclosures, while minimizing the risk of “hallucinations” by linking all such summaries back to specific SEC-filed notes. We are launching rDNA.ai initially with more than 2,000 high-profile deals involving more than 3,500 alliance partners. We intend to broaden our coverage even as we generate updated master summaries based on the latest SEC filings.

We intend to use this data to highlight trends in deal terms and alliance outcomes, much as we’ve done over several decades as Recap and then BioSci. We invite your feedback on whether this approach to tracking biopharma alliances is useful and reliable.