Funding Long Shots

by ryan_b1 min read28th Jan 20204 comments



This paper is about funding long shots, where long shots are defined as follows:

(1) low probabilities of success; (2) long gestation lags before any cash flows are realized; (3) large required up-front investments; and (4) very large payoffs (relative to initial investment) in the unlikely event of success.

They argue funding for these things is difficult because VCs are too small and institutional investors are not interested in long term high-risk, high-reward investments.

Their pitch is to create a new type of financial instrument, the Research Backed Obligation. These are securities which will be more appealing to regular investors, which they hope will expose long shot funding to a much bigger pool of capital. This matters to us because:

...society’s biggest challenges such as cancer, Alzheimer’s disease, global warming, and fossil-fuel depletion depend critically on the ability to undertake such investments.
4 comments, sorted by Highlighting new comments since Today at 11:42 PM
New Comment

This seems much the the argument behind mortgage backed securities. Some have good track records but others are be bit more shaky so seems a devil in the details type problem -- which then is multiplied when talking about "regular" investors, depending on just who that group is. Is the thinking here something like an ETF open to anyone with an investment account? Or a more restricted mechanism?

I also find the last blurb a bit concerning. I have seen in pointed out before that cancer research has been one to the most funded areas in medicine but producing results less the case. Could it be that we're making the error of thinking the problem is funding in a lot of these cases where as the real problem is one of insight -- which money is likely not to induce?

That said, I do see that there might be a place for this type of investment and it might be a good way to better formalize any number of crowd funding type of efforts that might otherwise go no where. However, packaging would seem to be a critical activity and might be worth a paper on its own issues and possible solutions.

Something like an ETF open to anyone with an investment account is the idea, but I have seen one of the authors pitch such a mechanism specifically for pharmaceuticals and in that case they said the resulting instruments would be prime targets for index and retirement funds. So I infer the target is regular institutional investors, including investment banks and hedge funds, rather than individuals like you or I.

This is definitely a devil in the details problem. Expanding on the specific pharmaceutical pitch I saw, for comparison with the cancer example: pharmaceuticals have also seen a lot of funding, but in fact there is less overall research being done; fewer drugs are even put forward through the process then previously. Most drug discoveries do not even start it. This is because only the ones that make it all the way through the process are profitable, so all of the development money has to go to the few best-chance drugs (as the company estimates it).

The way they say that their funding method solves this problem is because FDA approval has phases; so what they do instead is put a bunch of drug patents together in a bundle (like the mortgage backed securities), and then once that bundle goes through Phase 1 the survivors are re-bundled for Phase 2, and again for Phase 3. As a result, all the drugs get pushed as far as they can go. This makes a certain adjustment of the industry more feasible as well. Currently the pharmaceutical company must be the whole pipeline to capture any value, and it is hard to succeed as a research lab that just produces drug discoveries because any individual drug is so unlikely to pay out. Under the research backed security system they would be able to sell their discoveries into a bundle (or as a bundle) because there is a Phase 1 payout that is faster and more predictable. This is analogous to how mortgage companies do not collect mortgage payments, they create the mortgage and then sell the mortgage on.

Would be interested in more distillation of the idea here.

Can do. I thought about anchoring the idea on the much-more-familiar Y Combinator paradigm; do you think that would be helpful, or do you think it would be better to stick to a contained summary?