The Translational Mixer
The Translational Mixer
Episode 8: Andrew Lo on fixing business models in biotech and a sparkling Mio sake!
https://bit.ly/3YxRltJ. Andrew Lo, Charles E. & Susan T. Harris Professor of Finance and director of the Laboratory for Financial Engineering at the MIT Sloan School of Management, gives JC and Andy the skinny on his progress in finding new commercialization models for rare diseases. He also reveals plans on implementing his debt securitization megafund model, first described over a decade ago. Finally, he gives us some tips on how to prepare a delicious refreshing sparkling sake:
1) Mio sparkling sake
1 bottle Mio sake
1 freezer
DIRECTIONS:
From Andrew: "Store bottle of Mio at 58ºF (the recommended temperature of typical wine cellar). Move bottle to the freezer section of your refrigerator for about two hours before you're ready to consume it. The time is approximate, and depends on how cold your freezer is, so you may need to play around with this key parameter. After two hours, remove the bottle from the freezer; it should still be completely liquid inside (if it's partially frozen, you've kept it in too long and need to thaw it before opening).
Assuming that the sake is completely liquid, twist open the bottle and QUICKLY POUR OUT A SERVING INTO YOUR GLASS. The reason you have to pour quickly is that the sake is sparkling, which means the carbonation creates pressure in the bottle. Once this pressure is released, the liquid starts to foam and the foam will freeze, clogging the bottle and making it impossible to pour out any liquid. By pouring it quickly, the foam forms in the class, yielding the desired frozen treat. The remaining liquid in the bottle is blocked by the frozen foam in the neck, but this will melt while you enjoy the first pour, and should be ready to be poured out in just a few minutes (though the foam won't be quite as thick in the second glass).
This works best with Mio's individual 375ml bottles. It can be done with full 750ml bottles but it's hard to pour multiple glasses fast enough before the foam freezes in the neck (you have to line up your glasses and pour quickly in a straight line)."
Sources mentioned in the podcast:
Andrews new book:
Andrew W Lo and Shomesh E Chaudhuri. Healthcare Finance (Princeton University Press, Princeton, 2023)
Original paper describing megafund/securitization:
Fernandez, JM et al. Commercializing biomedical research through securitization techniques. Nat Biotechnol 30, 964–975 (2012). https://doi.org/10.1038/nbt.2374
Paper describing methodology for assessing risk (likelihood of approval) of biomedical assets:
Siah, KW et al. Predicting drug approvals: The Novartis data science and artificial intelligence challenge. Patterns 2, 100312 (August 13, 2021).
EU approval of Agilis Biotherapeutics/PTC’s Upstaza AAV2 gene therapy for delivering dopa decarboxylase to patients with aromatic L-amino acid decarboxylase (AADC) deficiency. As AAAC deficiency has an incidence of about 1 per 1,000,000 live newborns (332 potential US patients every year), a billion dollar return could theoretically be obtained in three years...in practice though the challenge for PTC is to find those patients!
The Mixer music “Pour Me Another” courtesy of Smooth Moves!
02:13. Financing ideas out of academia
07:58. The early-stage landscape for biotech
13:00. New models for rare diseases
15:03. Priority review vouchers
23:22. Funding ultrarare diseases
27:14. The future of securitization and megafunds
34:49. How to rate the risk of biomedical assets?
40:17. Who would invest in biomedical bonds?
43:43. The future of biofinance?
45:40. A sake and a daiquiri in one
Andy Marshall, Hello, welcome to The Mixer. I'm your host, Andy Marshall. I'm here with my friend Juan Carlos Lopez. How are you doing, JC?
Juan-Carlos Lopez: Hello, Andy. I'm doing very well. Very excited about this episode of the podcast. So who is our guest today?
Andy: So today our guest is Andrew Lowe. Andrew is the Charles E. and Susan T. Harris Professor of Finance and the Director of the Laboratory for Financial Engineering at the MIT Sloan School of Management. Andrew's research has spanned many aspects of economics over the years, but we're going to be talking to him today about his more recent work, which is this intersection between economics and healthcare. Another good reason to be talking to Andrew is he just published a book, Healthcare Finance, which he co -authored with Shomesh Childhury.
JC: This is great, Andy. I'm very excited about this conversation because you and I have admired Andrew's work for a long time, ever since we originally came across his first publication in Nature Biotechnology. It was one of those things that where you really say, yeah, this is a whole universe that we don't often think about. And we should because getting money for ventures is an important aspect of translational research. And he's had a lot of very innovative ideas about how to fund assets, how to fund emerging companies, how to maximize the number of projects that have an opportunity to get to the market. His thinking has had a lot of great influence in the field. And I am excited he accepted our invitation to join us.
Andy: Yes, I’m looking forward to it.
JC: Let’s get started
Andy. Let’s jump in.
----music---
02:13. Financing ideas out of academia
Andy: Andrew, we're really excited you could join us today. Welcome.
Andrew Lo: Thank you. Pleasure to be here.
Andy: I have your new book, Healthcare Finance, here with me. It's a tour of force, a voyage through the whole financing universe for healthcare. So, you know, as a way of starting, perhaps you could give us some background on why you've put this book together and your interest in recent years in focusing on financing and the sector of healthcare in particular.
Andrew: Sure, happy to do that. First, by way of background and disclosure, I am not a biomedical expert by any means. My training and my trade is as a financial economist and I got interested in healthcare about 15 years ago when a number of friends and family had to deal with different kinds of cancer. And it was through that process that I quickly realized that finance plays a really big role in drug development. I guess that shows you how naive I was. I used to think that there was a patient in need and there was technology that could help them, then money would magically appear to fund that technology all the way to FDA approval. And it wasn't until I started speaking to experts in the field that I realized just how piecemeal the whole funding process was, particularly for early stage drug development companies. And that's when I decided to take some of the ideas that I had been working on in the financial realm and apply them to biomedical innovation to see whether or not we can get more money to these drug developers so that ultimately these drugs actually reach patients.
Andy: It's interesting that you kind of mention how piecemeal funding is. Could you walk us through, let's say, a researcher has an asset who's got a paper in Nature, Science or Cell, they have some IP, and they're thinking this could be something that they could take forward. What are the kind of mechanisms that are engaged in kind of taking that forward through the translation process as you talk about in the book?
Andrew: When I first started talking to my colleagues at MIT and that out of the universities that were working on these kinds of early stage projects that would try to take ideas out of the laboratory and into the clinic, they would come to me for advice as to how to get funding. And being a finance professor, I figured that I should know something about it. And as I learned more about the field, it became clear that while there were a number of different sources of funding, how you approach them, when you approach them, with what materials do you approach them? All of these were shrouded in this layer of complexity and almost secrecy that, you know, there were initiated individuals who had been through the process. But for first -time entrepreneur founders, it just seemed really difficult. And the business models that were being deployed, namely venture capital, private equity, some nonprofit funding nowadays called venture philanthropy. All of those were really in various different stages of development back 15 years ago. Things are a bit better now, but over the last two and a half years, we've seen some of the same kinds of challenges, the so-called Valley of Death, that people refer to. So there are multiple programs out there that are attempting to fund early-stage translational medicine, but they are still relatively hard to come by. And there's no systematic approach that these first-time entrepreneur founders can use to be able to get the funding that they need.
Andy: There's the complexity of all of the different investment mechanisms that, frankly, befuddles a researcher if they're going through this for the first time.
Andrew: There's complexity, but there's also a cultural aspect. The scientist and the scientific culture is really not ideally designed to do fundraising. Scientists understand how to apply for grants, but how to raise money from investors, that's a different art form altogether. And unfortunately, I think that the culture of science, the very careful, reserved, measured type of explication that scientists engage in, that's really not ideally suited to investors. And it's not meant for engaging in flights of fancy, whereas when you make a pitch to investors, it's all about these flights of fancy, or as they would like to say, ‘vision’. And without that vision, it's very difficult for generalist investors to be able to deploy capital in this very important but nuanced to space.
Andy: I guess the other thing is there's also this volatility of the markets aspect to it as well, where you have these times when it's boom and then there's this time when it's bust, yeah?
Andrew: Yeah, Feast and Famine is one of the hallmarks of biotech funding. And that's exactly the worst kind of funding cycles that could be for biomedical development. It creates all sorts of perverse incentives, and it turns a scientific problem into a compound scientific and financing problem.
And those two problems coupled together can make it virtually impossible for early-stage, first -time entrepreneurs to be able to get the funding they need to launch their programs.
07:58. The early-stage landscape for biotech
JC: You see, this point that you raise, it resonates a lot with us because through our careers, we've interacted with a lot of scientists in the early stage who have exactly those limitations you point out. They are not particularly good at articulating a pitch or at trying to put in front of the investors what the value proposition might be. And fortunately, there are some investors who are very knowledgeable, and they would say, no, it doesn't matter. I get it, and I'm going to help you because I see something that you don't even see yourself. But what I've seen as well is that there are a lot of investors who really think in very short term, and they are much more concerned about, okay, but how far away are we from the market? Are we more than three years away from the market because then I'm really not interested. And of course, a lot of the academic research is more than three years away from the market. What I'm saying now is something that we've seen for 10, 12 years. But I wonder if you've seen any change in that aspect? Are there now more investors who are much more receptive to early stage? Or do you think that scientists are getting more money from things like SBIRs (small business innovation and research grants) or venture philanthropy, other mechanisms so that they come with something that is closer to market. Where is the movement happening?
Andrew: Well, as in most situations in finance and economics, there's good news and there's bad news. The good news is, you're absolutely right, there are more funding opportunities, both from the government, including various SBIR grants and other kinds of non-dilutive financing by venture philanthropists. As well, there are more investors out there that are looking to put money into healthcare. A good example of that is what happened during the COVID-19 pandemic. You had a situation where a large number of non-specialist investors came into the healthcare sector. And unfortunately, that was a bit of a feast and famine phenomenon as well because they came in at the early stages of COVID-19 in 2020 and 2021; and they pumped up the prices of a lot of different biotech companies, including many companies that should not have reached such lofty heights. And in the aftermath in 2022 and 2023, we saw a particularly vicious pullback and so for the last two and half years the biotech community has been living in a kind of a nuclear winter where it's very, very difficult to get funding. I do think that that's still good news in the end because NetNet, we do have more investors in this particular ecosystem than we had prior to 2019. So I think that that's the good news.
The not-so-good news, of course, is that the kind of volatility that we see in these funding cycles still makes it challenging. And what makes all of this worse is that the pace of scientific innovation has increased. So even though it might be the case that there are more investors out there, and net net, there may be more dollars that are available to entrepreneur founders, the fact is that there's so many more ideas that are being developed—you know, gene editing, gene therapies, cellular therapies, mRNA, siRNA—all of these new modalities need funding. And so the pace of scientific innovation has outstripped our ability to take advantage of these things to ease the burden of disease. So even though there is more money, I feel that we are still in a situation where there are some great shortages and there’s so much more good that we can do to change to bring more resources to bear to this important industry.
Andy: It seems like there’s two kind of layers on top of this. One relates to the fact that these conventional models they do the entire discovery and development process for one assets, or maybe one or two assets, they build it one at a time. So you keep on having to build all this one-time infrastructure and it increases the cost of development I guess?
Andrew: That’s right. The analogy that I came across a number of years ago where I was first first learning the process of drug development is: imagine a soccer match where for every single match you build a stadium, you fill the seats with it in order to watch this one match, and as soon as the match is over, you dismantle the stadium, and you wait for the next match, and when the next match comes along, you build another stadium, and the process goes all over again. That really does feel very much like a clinical trial. There has to be a better way, but of course its very difficult to innovate in something that is so expensive and where conservatism is the the hallmark of good scientific progress.
13:00. New models for rare diseases
Andy: The other layer, in addition to that, is where our last guest actually was in the area of diagnostics. For instance, in the area of diagnostics, this kind of venture model just doesn't seem to work very well. And then we're also very interested in rare, ultra -rare and nano -rare disease as well, which I guess is another sector of healthcare that is poorly served by these conventional models, Andrew, yeah?
Andrew: Indeed. Although I think that with some business innovation, it is possible to address these types of challenges. For example, in the case of rare diseases, a few years ago, some of my colleagues and I started thinking about using portfolio theory to de-risk the portfolio of biomedical programs. So instead of going after one disease at a time, if you went after multiple indications and created a pool of investments, you could actually attract more capital, reduce the risk. And those ideas ultimately were the basis of a company Bridge Biopharma that I co-founded in 2018 with Neil Kumar and others. And that's done reasonably well in terms of being able to attract funding, but also in developing therapies for rare diseases that were currently not being dealt with by big pharma.
In the case of ultra rare, I think that even there you can make progress particularly with government programs like the priority review voucher (PRV) which of course is up for renewal and I’m hoping it will get through US Congress. That’s an example where the government has provided some pretty useful incentives for drug developers to focus on ultrarare.
Nanorare, where we're talking about a handful of patients around the world, that is really a challenge. There are some non-profit groups like n-lorem that are making great progress there, but those really depend on the generosity of donors to fund. Whereas I think that if we are able to pool all of these conditions together into a larger portfolio, that it's possible to fund all of them.
15:03. Priority review vouchers
JC: I was very interested in a lecture I heard you give not so long ago where you talk about a potential model for ultrarare that depended upon priority review vouchers in which one of your slides was this table in which you indicated the number of programs you needed, and how many needed to be successful and the fact that probability of success was much higher in a portfolio of these programs. And I thought that that was pretty inspiring and the math makes sense. But I wonder, do those numbers you came up with for that exercise, do they include things like manufacturing and distribution? Because I think that that's one of the big challenges for ultrarare. For rare disease, you can see how people can manufacture and distribute worldwide. But for ultrarare, where the patient are spread across the globe, is the model geared more towards the American market or did you make any assumptions there that you didn't have time to elaborate in the lecture?
Andrew: You know, as usual, as an economist, you make all sorts of assumptions and sweep things under the rug. So guilty is charged. But you're right that the model that I was implicitly focusing on was the US model where you're able to charge relatively high prices because you're delivering transformative therapies. In other countries, particularly single payer systems there’s much more of a constraint on how much you can get payers to cover. But even having said that the key is being able to pool these projects into a larger portfolio because that reduces the risk of the overall investment.
So I would say that the combination of being able to charge reasonable prices, being able to do this on a broader scale across multiple indications at the same time, and also being able to have the technical expertise to have a reasonably high probability of success and also to be able to deliver these drugs if approved using the proper GMP (good manufacturing practice) facilities and so on, I think that those elements put together will make these drugs viable. That was the surprising thing. Even with ultra-rare conditions, having programs like the US PRV can actually incentivize drug developers around the world to come to the US to develop these drugs, and then once they're developed here, they can be distributed and deployed worldwide.
So I think that it is an example of a business model that I think will work, but it does contain some assumptions that need to be more carefully vetted across the various different venues around the world. For example, in certain countries where they have a very severe cap on pricing, this model will not work. It will not be sustainable. So you really have to do the math.
And the interesting thing is that in this calculation that I did on what the math looks like. There are three components. One is the cost of drug development. Two is the probability of success, which is based upon the science and medicine. And three, it's the cost of the PRV or the ability to charge a certain amount of the drug.
So in one equation, you've got the science, the medicine, and the financing all rolled into and they are interdependent. You can't decide on one without looking at the other two. So I think that the formula that I’ve put together is designed to be used by a variety of different stakeholders to develop these important therapies
Andy: How certain can you be though that priority review voucher legislation would be reauthorized. I mean, it seems like a big question mark, Andrew?
Andrew: I agree. And anytime you deal with lawmakers and policymakers, it's really tough to handicap because, as we've seen before, anything can happen. Having said that, I think that among the various different programs that are out there, the PRV is probably among the least controversial simply because it affects a relatively small population of patients that are, in many cases, dealing with terminal illnesses, and they have no other course of treatment. So from the perspective of trying to do something good for society, it's kind of hard to argue against the PRV. The other thing is that in many of these cases, because of the treatment of these ultra -rare conditions, you're talking about really tremendous value added to patients. In some cases, turning a death sentence into a completely manageable chronic condition or a cure altogether.
So I think as far as different programs go, I would think that the PRV is among the most attractive from a political point of view. And there are much bigger targets that politicians have in their sites. For example, the price of insulin. That's something that, you know, we've had insulin for decades. Why are we still paying the crazy prices for that we are? I think those are bigger targets for politicians to take on.
JC: Yeah.
Andy: Another question about that model to get the maths to work. As you mentioned, is development cost, yeah? I've heard you use the example of $10 million for development and that gives you a return of 20% with a PRV. But how certain are you that you could put together—and I guess this fits into ideas you have about a diversified portfolio within a single entity. It's like by doing that, you can start to hammer down some of these clinical trial costs. But it seems to me, Andrew, like a constant battle, yeah, because all the time it seems like clinical trial costs are increasing and increasing. What's your thinking about the reality of where one ends up with a cost of a clinical development program?
Andrew: Well, the reason I wrote the paper and put out that equation that related the cost of development and the probability of success and the rate of return on a particular project, is because I wanted to give scientists an understanding of what they need to do in order to take an idea that they have and put it into a patient and save a life. Because one of the things I’ve learnt from spending a bit of time with biomedical experts is that they are incredibly creative, but they need a goal. So if the goal is to get a drug to approval without worry about what the cost will be—and many people argue that is what big pharma does—they spend enormous amounts of money in order to get a drug over the finish line. That will lead to one set of behaviors.
But if you talk to scientists and let them know that if they're able to lower the cost of drug development to a certain level, that's going to mean the difference between life and death for patients that currently don't have access to that kind of a treatment. That leads to a very different behavior. I mean, I've seen examples of scientists who have engaged in all sorts of innovation in order to be able to take a fixed pot of money and squeeze out the most incredible science out of it because that's what they were given and that's what they were told they needed to do in order to make their breakthroughs. I think that scientists aren't given enough credit for being able to think economically because even though they don't have training, they certainly don't lack quantitative skills.
And so if you just explain to them that by getting the cost of drug development down from, say, $50 million to $10 million, you can actually make a whole series of diseases go away. You can cure a number of diseases if you're able to bring the cost down and if you're able to increase the probability of success. If you just tell them, that's the goal. I things will happen.
23:22. Funding ultrarare diseases
JC: I mean, there are a lot of people who are beginning to play in this space. And I think that actually a lot of the innovation in terms of coming up with new models to try to develop therapeutics for small populations of patients come from disease foundations and even from single parents. There's a lot of energy in that space. I find it very interesting because, of course, there are the nanorare patients in which there may be only a few, right? Let's say 30. And there's a model for that. You mentioned in n-Lorem and they're doing great stuff. And then there are diseases that affect, say, a thousand people, right? And then for those, a lot of what you're saying applies to that kind of positions, and that may be attractive to a company. But there are a lot of diseases that fall in between, right, between 30 and a thousand. And those are the ones where I think there's a lot of energy and where there's a lot of new approaches that people are trying. And some of them are actually either inspired by what you're saying or they stumble upon these portfolio ideas. What I've seen so far is that some essentially non-for-profit biotechs are putting together portfolios of, say, five projects and are assuming a probability of success of 40% so that two of them work. And then with that, if they get one success they can pay for the five existing programs, if they get two successes they can take on five new programs, and that sounds perfectly fine on paper. But what we need right now is a win. So I’m very curious…
Andy: Is that using priority review vouchers JC?
JC: Yes exactly. Yes because there’s only a population of a fewer than a thousand people, right? So you're not going to get any significant money from sales. So it's mostly through the PRVs. They say, okay, if we get two approvals out of these five, then we have $200 million on the back of PRVs. And with that, we can pay for the clinical development of these five programs and then five more programs. But some of the assumptions that they make are pretty interesting. They have to do with being able to do short trials and measure surrogate markers. They are assuming also that they can do a pivotal trial with just 10 patients, which may be. If there are only 100 people in the world with the condition, I guess 10 is a lot. But I think that these assumptions need to be pressure tested. And once we have a win, then we'll be able to see how successful the model really is.
Andrew: Absolutely. But I think that in a way it has already been pressure tested, and we have seen some successes. For example, in the case of a company that I wrote a case study about, Agilis Biotherapeutics, they were developing gene therapy for a very rare condition called aromatic l amino acid decarboxylase deficiency. It's a neurological disorder, and I think with 10 patients, they got an FDA approval, and PTC acquired Agilis, and now the drug is approved in Europe, and I believe that it's being considered by the FDA this year.
I think worldwide they probably don't have more than a couple of hundred patients, but yet that was a very profitable investment for a number of investors. So I think we're going to see more examples of that in the case of gene therapies, gene editing, ASOs possibly, mRNA, a number of different drug modalities that are targeting these ultrarare diseases where the efficacy is just enormous. It provides years of quality adjusted life. I think we are living in a golden age of biomedical innovation, but we also need to support that with a similar golden age of financial innovation.
27:14. The future of securitization and megafunds
JC: You know, another thing that we've been very interested from reading your papers and hearing you talk, and this is returning to portfolio theory and having large numbers of projects to spread risk and reduce the Sharpe ratio, and all these things I've learned from listening to your talks.
Andrew: Thank you.
JC: We are quite interested in your original idea that you published in Nature Biotechnology of the megafund, which a lot of that has come to pass, and it's quite commendable that you set the foundation for a lot of that work. But one of the things that we haven't seen being too adopted is the securitization of this debt. And we're wondering if now, with 10 years after that paper was published, maybe more. If you have an understanding of why is it that that part has been a tougher sell. Portfolio diversification, I think people get it, people embrace it, and we've seen it. But the idea of raising a megafund and doing securitization doesn't seem to have been as successful, and we wonder if you have any thoughts about why that might be?
Andy: It might also be helpful, Andrew, because a lot of people find this quite bamboozling just to kind of lay out exactly what the idea is.
JC: Yeah, what I'm talking about. Clarify what I'm talking about,
Andy: yeah.
Andrew: Well, so the very first paper that I ever published in the field that I now call healthcare finance was in 2012. It's a particularly meaningful milestone for a couple of reasons. One is I had never heard of this journal called Nature Biotech until 2012. And I just asked some of my colleagues in MIT, where would be a good place to publish this kind of a crazy idea of applying portfolio theory on securitization to healthcare assets?
And they said, oh, well, you ought to try Nature Biotech because all of the biotech VCs and the scientists and entrepreneur founders, they read that journal. So not knowing any better, I submitted a paper and it was a long and very important process because I learned how difficult it is to make sure that you get all of the narrative exactly right for this audience. But it was a very, very important publication because it raised this kind of a possibility to an audience that really was key to taking some of the ideas implementing it.
Fast forward, it's been now 12 years, and we've seen a lot of these ideas implemented, but the one idea that has not yet been implemented is the notion of using securitized debt to finance biomedical projects. And let me explain what that is. Portfolio theory, everybody gets. Multiple shots on goal, reduces the risk, increases the likelihood of success. And that's been implemented by a number of biotech and pharma companies since then.
But the second element that I proposed was once you reduce that risk, then instead of financing using the traditional sources of capital—venture capital, private equity, public markets with an IPO and so on—instead of those traditional sources of financing, why not borrow money? Financing by issuing bonds. This is something that is done all the time in the financial industry. And one example of that that originally got me thinking along these lines was the financial crisis, which was caused by a huge influx of these kinds of securitized debt instruments to provide mortgaged-back securities. And the idea during the financial crisis was put together a lot of mortgages into a portfolio and then raise money from investors looking to finance home ownership.
And that's really what ultimately led to the financial crisis. Most people therefore were skeptical about using these techniques to fund biomedical innovation because, well, didn't it cause the financial crisis? And I have to say that that's exactly why I thought we should use these methods. You see the financial crisis happened not because these techniques didn’t work. It happened because these techniques worked way too well. They were powerful tools f or raising large amounts of money for a dedicated purpose. And the same thing can be done for cancer. So why hasn’t it happened?
There actually is very simple reason, which I've been working on, particularly over the course of the last few years, the reason is this. In order for people to be comfortable buying bonds, essentially lending money to various different counterparties, is they need to know the risk. They need to know what the probability of default is going to be. And so we know that in the bond market, we've got rating agencies that can rate bonds as AAA, very safe, AA, safe, but not all that safe, or BBB, which is really, really unsafe and likely to default. How do rating agencies rate biomedical projects?
That was missing 12 years ago. And what I've been doing over that time is thinking about ways of developing mathematical models to forecast the probability of default of bonds backed by biomedical projects. This is basically tied to forecasting probability of success of a clinical trial. And a few years ago I published some papers on using machine learning techniques to forecast clinical trial outcomes. And have now been in the process of developing a commercial model of default probabilities so as to be able to allow investors to basically handicap the likelihood that they'll get repaid if they decide to buy a bond in this area.
My sense is that within the next year or two, we will finally have securitized debt where the underlying collateral is a combination of these biomedical projects. So it's taken a while, but it's going to be happening within the next year or two.
JC: That's very interesting. Who would issue the debt, the company that's developing the therapeutics, or who is responsible for paying that?
Andrew: Typically, the entity that's issuing the debt is the asset manager that is in charge of funding all of these different clinical trials. So, for example, imagine a pharma company that wants to fund lots of different clinical trials. They may decide to issue the debt and use the money that they get to fund those clinical trials. And if for whatever reason they can't pay the interest on the debt, then the bondholders take control of the assets. But it doesn't have to be a biopharma company. It could be a financial portfolio manager whose job it will be to dole out money to the various different biotech companies that are engaging in the clinical trials. And in exchange for providing those biotech companies with the funding, this portfolio manager will receive royalty interests in the drugs that are being developed. So it could be a variety of entities.
It can actually even be a nonprofit organization that decides to borrow money from for-profit investors to be invested in a particular disease like cystic fibrosis or Alzheimer's or Parkinson's. And the nonprofit organization basically invests in these various different therapeutics with the idea that the royalties that they receive will eventually be able to pay the interest for the bonds that have been sold.
34:49. How to rate the risk of biomedical assets?
Andy: I think that's fascinating as well, Andrew. The question on my mind, though, is very often in biotech, we're going after conditions that are entirely new territory or slightly a step aside from an indication that has been explored before commercially. So, I mean, we were just talking about rare diseases where there's going to be very little commercial precedent in terms of activity in those diseases. So past performance and the way in which that's used to train these machine learning models sounds intriguing, but I would worry about the incompleteness of understanding of biological models and the incompleteness of overlap of each case where you're developing a therapeutic in a new area of biology and how such a model would give you an accurate rating for what's the level of clinical attrition one would see.
Andrew: So that's a really good point, and I'll answer it a couple of different ways. The first answer is that even though we're talking about predicting, in many cases, completely novel therapeutics, the fact is that the success for failure of a clinical trial is based upon a number of factors, some of which are unprecedented, but most of which are actually not unprecedented. And in fact, many of them are pretty well known and understood. So one good example is team/manager track record. If a biotech company has had a single drug approval prior to the one that you're handicapping, it turns out that that one approval dramatically increases its likelihood of success. Why? Simply because it had that experience. So that's an example. Whether you're developing a completely novel therapy or if you're developing a Me-Too cancer drug, in both of those situations, sponsor track record is important. I would argue that the vast majority of the factors that allow you to predict success or failure of a clinical trial are things that are well known and where we have a fair amount of data. And therefore, these machine learning models are actually reasonably accurate, even for completely new therapies in completely novel applications. So that's one answer.
But the other answer that I think may be even more relevant is that the level of scientific precision that is needed in order to get a nature biotech publication is completely different from the level of accuracy that an investor needs in order to put capital to work. You know, when an investor puts his or her life savings into the S&P 500 mutual fund, there is no guarantee that they're going to be making any kind of a return. They could lose 20% in a given year. It has happened with the S&P 500. But there are also years where it has increased by 30%. And the kind of randomness that goes on in financial markets, I suspect most scientists don't fully appreciate just how noisy the investment opportunities are that investors have to decide upon every day.
So I would argue that unlike scientific progress, unlike the drug approval process, the accuracy that you need to make a financial investment is far, far less than you would in other scientific endeavors. And therefore, even though you can't handicap perfectly accurately, you can actually estimate to the same or better degree of accuracy than financial market participants expect right now with their bonds in other industries.
That's really the standard of care in the financial realm.
Andy: Another way of asking the question that I just ask is perhaps, given that there's this precedent data, presumably there's a higher likelihood if you're ticking more of the features box, whatever it is in the model, that look like development will be, have a better outcome, then these will be in the highest tranche, and then it will go down. As you get more and more risky, it will go into junior trenches or whatever.
Andrew: Exactly. The whole point of financial markets and different types of financial securities. The whole point is really to provide investors with the right level of investment opportunity that they're seeking. By level, I mean the right rate of return that they're looking for, but more importantly, the right level of risk. So, for example, the really risky investment opportunities in this world, they go to hedge funds and private-equity investors. The really safe assets in the world, they go to banks, mutual funds and retirees that literally cannot afford to lose any of their principal. So by following the dictum of giving investors what they want, we can actually raise a large amount of money by creating the financial structures to support biomedical innovation and splitting up the risks to give hedge fund managers the biggest risks, the mutual fund managers, the lower risks, and the banks and the money market funds, the least amount of risk altogether. That's the basic idea behind securitization.
40:17. Who would invest in biomedical bonds?
JC: See, this reminds me of another thing that I heard that one of the, the first time that I, that I heard some of these ideas was at one of your or Cancer-X conferences. And I remember the first time that the whole concept was presented, one of the points that people made, and I don't remember if it was you or someone else, was about the fact that every household in the US chipped in into the megafund, everybody would need to chip in. I don't remember if it was $8 or $18. It was a very modest amount, but that's not what you wanted, that you actually wanted professional investors coming in with the cash as opposed to retirees or these kind of people. And I wonder if your thinking has changed on that front, or if you still think if its better to stick with professional investors or if we need to go to every household in America?
Andrew: I think that it has to come in stages. I think that first, professional institutional portfolio managers. They should be the ones leading the charge for making these investments. Mainly because they’re new, they’re risky, nobody has any experience with them, and therefore you really want to have professionals taking on those risks. However, once that initial investment is made by these institutional investors, once the market develops expertise in these investments, at that point we can then start to develop mutual funds and ETFs focused on different diseases. I mean can you imagine an oncology focused ETF? How many people in this country or around the world have lost friends and family to cancer and don’t know what to do about it? They want to help but they don’t know how to invest in venture capital or private equity. But imagine if you started an oncology ETF that focusing on the most difficult types of cancer using transformative therapeutics to make these kinds of breakthroughs. Everybody who has been touched by cancer would like to invest $10, $15, $20, maybe $1,000. Imagine if that were to happen, you’d get tens of billions of dollars almost overnight. The issue with that is that if these individuals end up getting burned that could set the industry back by a decade. And so I want to be careful about how these financial innovations are developed. Once we see that these risks are manageable and that you can actually get institutional investors to understand and explain these risks to the retail investors, at that point the retail products can come, just like right now we're starting to get interesting retail products in crypto. I'm still not a big fan because I think that crypto is a very risky investment that most retail investors don't really understand, but you're getting more retail investors that are comfortable at these kinds of risks now than they were 10 years ago. So the same thing I think could happen with healthcare-based products.
Andy: Yeah, we should clarify what ETF means for our uninitiated audience. That's a kind of exchange traded funds, is that right?
Andrew: That's right. Yes, it's an exchange traded fund, which is kind of like a mutual fund where you have a basket of securities, but it differs from a mutual fund in that it trades all day long just like a stock. So it's like a one stock company, but it is a representative of multiple assets within that portfolio.
43:43. The future of biofinance?
Andy: One thing that would be great, Andrew, is if you could give us your vision of how you see this landcape developing over the next few years?
Andrew: My dream is to be able to see many many more financial models to deal with the various kinds of therapeutics applied to different patient populations so that no patient is left behind. Obviously, the big pharma companies are going to continue focusing on the blockbuster indications, but we now have scientific progress being made almost on a daily basis where we literally have the ability to cure life-threatening diseases for small populations of patients. I think it's unconscionable that we don't take those ideas and bring them to patients. And so I'd like to see large amounts of money on the order of multiple hundreds of billions of dollars coming into this industry through these new financial methods to fund therapeutics.
Because ultimately—we all say this—because if you don’t have your health, you don’t have anything. Well if that’s the case, then we should be investing most of our money in our health. And right now we’re paying about 17% of GDP for healthcare. And people say, well, how high can it go? How high would you be willing to spend your income in order to save your child or a family member from devastating disease? 30%, 50%, I'd be willing to spend 120% of my income to be able to save my kids. If we think along those lines, that means that there's a lot more capacity for our economy to be devoted to healthcare. Because if you don't have your health, you don't have anything. So why bother saving money and spending it on things that don't provide lasting value, whereas we can actually create life by investing in the therapeutics for all of these terrible diseases?
45:40. A sake and a daiquiri in one
JC: Thank you, Andrew. That's a very inspiring thought, and I certainly share it completely. So, as we may have mentioned, the podcast is called The Mixer, because we try to mix our interest in translational medicine with our interest in cocktail making. So we always ask our guests what is their go-to drink, when you have some downtime and you want to think about the next book you’re going to write, what is your go-to cocktail or your go-to drink.
Andrew: Well that’s a very difficult question because there’s lots of drinks that I enjoy. But I do have one that I think is kind of unusual; I dare say that none of your listeners will have ever come across this because it's not actually a mixed drink. But it is a cocktail in that it requires a certain process to create it.
So I'm fond of Japanese sake's, particularly sparkling sake. They tend to be quite sweet and fruity. Not for everybody, but I enjoy it. But there's a particular way that I enjoy the sake. And the brand that I favor is a label called Mio, M -I -O. It comes in a dark blue bottle, beautiful and delicious, refreshing, crisp, sparkling sake. But the way that I enjoy it is I keep my wines at 58 degrees F, including the sake. And when I want to drink it, I'll take it out from the wine room and put it into my freezer of my refrigerator. And I will leave it there for exactly two hours. And it's important to get the timing right.
Now, obviously, different refrigerators may have different temperatures. I'm just telling you about my refrigerator. You put it in the freezer for two hours, and you take it out, and it's still liquid. You open the twist cap and pour it out very quickly. And what happens is that it starts to foam and the foam freezes. And what you get is a kind of a frozen daquiri-like sparkling sake that is amazing.
Andy: Wow!
Andrew: So I would urge all of you to try this. It's a very special recipe, and sommoliers would, I'm sure, pale and be horrified by this, but it is an extraordinary drink, and I hope your listeners will enjoy it.
JC: I don't know about our listeners, but I'm certainly going to give it a try.
Andrew: Me too.
JC: I'm going to give it to try. And, you know, normally we put the recipe for the cocktail as part of the podcast. This time, we'll type these instructions, and will even provide a link to this sake. It was very good intelligence Andrew. Thank you.
Andrew: My pleasure. Thank you so much. Its been a honor to be on your podcast.
JC: Its been our pleasure. Thank you for your time and insights Andrew
Andy: Bye Bye
---- Music---
Andy: So that was a really great conversation with Andrew, I think. He's one of these individuals who really thinks creatively about ways in which we can innovate in financing different types of biomedical research. And it was interesting, I thought, JC, that he started to really outline how he's thinking about putting into practice the securitization model, yeah?
JC: Absolutely. I've always been a great fan of his work. He, as you say, is a very creative individual. And certainly that was the part that I thought was most novel of this conversation. The fact that he hasn't given up on the idea of issuing bonds and how issuing this kind of debt could lead to unlocking a lot of capital that is currently unavailable.
So I'm really looking forward to see how that plays out. Of course, the devil will be in the details because how well can you really rate biomedical assets?
Andy: Right.
JC: Right, that's the challenge. He is certainly the right person to be thinking about this because he has been thinking about this space extensively. So I'm really eager to see the results of this latest line of inquiry that he's launched. He always delivers, so I'm confident that it will be a very stimulating contribution yet again.
Andy: Yeah, he outlined this idea of using these machine learning models to give you a level of risk associated with a particular asset in a particular indication, yeah? And as you said the Devil will be in the details. Andrew brought up the fact that it doesn’t have to be as precise as you would want predictions to be in early-stage biology and understanding mechanisms and so on and so forth. But I think it is a real challenge yeah? Assessing risk of assets in different indications becomes easier the later the asset is in clinical development. So perhaps this starts later in clinical development than in early development where there are so many other factors that are really difficult to really, I would argue, train a model to give you an accurate output and level of confidence in the prediction that the model's coming out with, yeah?
JC: Absolutely. There are too many variables. Also the volume of data that you have to train these models. Doesn’t strike me that there is that there is the same amount of data that you have in other applications. But I could be wrong. Again Andrew would know better than us. And also it did sound like he is focused on late-stage assets. On clinical trial outcomes and these sort of things that are important for taking drugs through the last steps of the process. I wonder how successful the rating will be in early stage assets. That’s something we didn’t have a chance to discuss but maybe it can be something we can explore in a subsequent conversation once he publishes his current work.
Andy: The other things that I still think we are left somewhat up in the air about is these indications that currently lie outside of where conventional biotech financing works, yeah? I mean, we spent quite a bit of time in the conversation talking about ultrarare and nanorare and making the mathematics, the equations work kind of really seems to revolve around how much money you have to spend on clinical development and weighing that against a $100 million ticket price for a priority review voucher if you get one of these things approved. It's a continuing area of interest for me understanding how you push back down on clinical development costs.
JC: Right. That's one of the challenges, the cost of development. To me, as I said during the recording, there are different people who are already trying to apply portfolio theory to these kind of assets. And what I've seen is that everybody is quite happy with getting the drugs through the approval process, which is great. That's when you get the PRV and that's when financially it has made sense because you have made some money on the back of a relatively small clinical trial. But then what?
You have this asset that was approved. How do you get this to patients? That doesn't seem to be the focus of a lot of these models, right? And there are ways. You could say, okay, so if I am a charity and I get the PRV, I'll spend all of that cash in delivering this to kids. I'll fly kids to centers of excellence so that they receive the treatment.
Andy: Right.
JC: But if you are somebody who wants to make a return on investment, that's certainly not the way that you want to do it. So then what do you do? You just manufacture a few? You put them out in a center of excellence, and only people who can be flown there or who can raise the cash to be treated, will go and take the therapeutic? That is the part of the process that I find unsatisfactory. And this is no criticism of what Andrew is doing, right? I mean, that's not the purview of his work. But to me, it's something that I haven't really seen addressed in a satisfactory way.
Andy: I couldn't agree more, but I think this is a kind of area where we just have to dig in more and when you have creative minds like Andrew thinking about this, I am kind of optimistic.
JC: I am too
Andy: Talking about creativity, what did you think about his choice of beverage?
JC: Yes very original. I am not really an expert on sake by any means. Its not something that I appreciate as much as I should. I know that there are very high-quality sakes in the world. Unfortunately, I've never really developed the taste or the expertise, but that doesn't mean that I'm not going to go immediately and buy a case of these sparkling sake and do a time curve to see if my freezer gives the expected result.
Andy: Yeah. I'll be the other research site.
JC: Exactly. We start with an hour and a half, an hour 45 minutes, two hours, et cetera, until we get the biophysical event he described. So I'm going to go and make that investment as soon as we sign off, Andy.
Andy: Sounds great. So we'll put all of that information down in the description below the podcast. And we look forward to you joining us next time. You can follow us on Twitter. I think it's @themixerpod is our handle so you can get updates on newly released episodes and we look forward to seeing you next time.
JC: Always a pleasure to be here, Andy. Take care and see you next time.
Andy: Cheers, JC. Cheers everybody.
JC: Cheers.
