# New Funding Models for Biomedical Innovation with Andrew Lo | Markus Academy | Ep. 51

https://www.youtube.com/watch?v=223uT_FJ36Q

[00:04] welcome back everybody to another webinar organized by princeton for everyone worldwide.
[00:09] we're very happy to have andy lo with us from mit.
[00:12] hi andy.
[00:12] hi andy will talk about new funding models for biomedical innovations.
[00:20] so we're very excited to learn more and i think especially with the kobe crisis this became very important and very prominent.
[00:25] so we will learn more about this today.
[00:27] so let me give you a few opening remarks about the innovation more generally.
[00:35] so as we all know innovation ideas are non-rival goods.
[00:41] so you can just share it with many many without costing anybody who shares it that's very very special about the ideas and innovation.
[00:48] and polygroma always refers to this water bottle.
[00:50] the water bottle itself is a regular good.
[00:54] but if you have an idea how to you know mix it with some certain salt and water and fight diarrhea with that.
[01:01] the idea how to fight some medical diseases.
[01:05] in particular here diarrhea is actually free you can just spread it around without causing anybody for spreading.
[01:13] so that's one big aspect of ideas.
[01:16] the other one is you know r d expenditures they have positive externalities and hence there's always some underinvestment.
[01:24] so what we want to do today we want to understand this uh better and there's of course an interplay between the innovator the entrepreneur the government the state and the venture capitalists or venture funding or the capitalist and system and there's an interplay between these three entities.
[01:40] you know this triangle i drew from bill chainway's book which you know studies innovation in the economy in much more detail than what i can mention here briefly today.
[01:51] so what what's innovation what's interaction with the government and we will see i will focus here a little bit on the government and i think we'll focus much more on the venture capitalist financing aspects.
[02:01] so basic research of course is uh subsidized by the government because very has very long horizons and high
[02:07] risk investments you don't know.
[02:09] the the explicit outcome but how can the government typically subsidize and help to get innovation going given that the positive externalities and there's under investment?
[02:19] one is just to subsidize the investment itself the cost.
[02:23] the other one is just to buy the end product to have some demand pull.
[02:27] and and by that and the third way is to ground patents and you know to give essentially temporary monopoly rights or allow for some price discrimination.
[02:37] but of course you know withholding monopolists withhold certain products from the poor and that seems a little bit immoral in terms of the health consideration so one has to probably be very careful with this aspect and a fourth way is to provide a price you know we had an earlier webinar last april or may where michael kramer was presenting and we talked about the x prize and price uh where the government provides a price for certain inventions.
[03:05] in general i think it's very important.
[03:07] that the government absorbs some of the risk
[03:10] and they risk essentially innovation and co-invests and one way to co-invest essentially is to uh you know charge taxes
[03:19] if you have low profits you pay low taxes
[03:21] if you have high profits you have high taxes
[03:26] so what are the two main innovation models maps will come to this
[03:29] so the traditional innovation model i would say for continental europe and longer tradition essentially of this large corporations and within the large corporations there's a lot of r d activity and that allows that the large corporations can try different ways and share the risk of different approaches
[03:50] you know whether one approach might work out the other approach might not work out but this actually was now supplanted by a startup model
[03:58] where actually most of the r d is done by little startups
[04:02] and then the large firms buy the successful startups
[04:04] so it's more a takeover so there will be
[04:08] takeovers of startups.
[04:10] and that was very prominent in the pharmaceutical industry in particular.
[04:13] where you buy startups and then the big companies have distribution networks to market.
[04:18] and then test the products and place it in the marketplace.
[04:22] and it's interesting to see how the world has changed towards the startup model and of course the venture capital funding for the startups is very very crucial for this as well.
[04:34] but in general we would like to have a resilient society so i'm always coming back to this resilience term and development very be very agile.
[04:42] develop new medical uh innovations and vaccines in particular.
[04:47] and for this we need a very good venture capitalist ecosystem and and have optimal risk sharing.
[04:54] and here comes another aspect of our capitalist system is you know to what extent do we welcome failure and to what extent do bankruptcy lead to some stigma.
[05:04] and if you go back in the old principles of other liberalism there's this liability principle which says you know.
[05:10] whenever you do something you're liable for it and there's you know 100 percent liability you can't get out there's no bankruptcy protection and there's of course this limited liability principle and depending how strict you put this limited liability aspects to it you allow for more risk-taking and less risk-taking and perhaps you know for this innovative aspects you might to promote it is sticking here because of the social benefits and you know that's where the bankruptcy code plays an important role and as i mentioned venture capitalism plays an important role because most of these innovations are real options so the optimal risk sharing and the real option component is very important and the venture capitalists also provide a lot of advice and expertise as they go across various different startups and bring expertise from one startup to the next startup and that's an important role the venture capitalists do on top of funding the projects so with this little introductory remarks i will pass on the floor to andy who knows way more than i do in particular
[06:11] and he will talk about the new funding models for biomedical innovations or focusing on biomedical aspects and you know he's worked for many years on in this space and he also teaches a course on health economics and health finance at mit so drawing from all his insights we get now compressed view about his uh what he's learned over the years in here so thanks a lot andy it's great to have you with us and uh the floor is yours.
[06:41] okay terrific thanks so much marcus um can everybody hear and see me okay.
[06:46] all right great so i i want to start by thanking marcus and the benheim center for finance at princeton for organizing this great event and for giving me the pleasure and the honor of speaking to all of you today.
[07:00] and i want to thank all of you for joining from around the world i hope you're all staying safe and healthy amidst this pandemic and very much appreciate your being part of this.
[07:08] so as marcus mentioned i'm going to be focusing today.
[07:12] on a topic involving healthcare finance
[07:14] and funding biomedical innovation
[07:17] and i have to start with a disclaimer
[07:19] and my disclaimer is that i am not
[07:21] a biomedical expert by any means and in
[07:24] fact i'm not even a healthcare economist
[07:26] i'm i'm a financial economist by
[07:28] training and trade
[07:30] much like marcus and in fact i was very
[07:33] interested to
[07:35] hear about his research on resilience
[07:37] because
[07:38] this is exactly the point of why we need
[07:41] a robust health care system
[07:42] and it's not surprising that given
[07:44] marcus's path-breaking research on
[07:46] systemic risk
[07:47] i think it naturally led him to start
[07:49] thinking about what it makes
[07:50] what it takes to make an economy that is
[07:53] more robust and and resilience is an
[07:56] important aspect i'll be talking about
[07:57] that
[07:58] in the very specific context of of
[08:00] healthcare
[08:01] so as i said i'm not an expert in
[08:03] healthcare finance why why did i
[08:05] start getting involved in this and it
[08:07] really has to do with personal reasons
[08:10] a number of years ago friends and family
[08:11] were dealing with various kinds of
[08:13] cancer.
[08:13] and i have to say that i was really taken aback by that.
[08:17] i had really never dealt with these kinds of issues myself and and even my close friends and family members were spared until about 10 years ago.
[08:26] when in the space of four years six people close to me, including my mother, all died of various kinds of cancer.
[08:33] and that was a pretty big wake-up call for me.
[08:35] during that process i tried to do what i could to be sympathetic and helpful to to my friends and family.
[08:41] and so i started reading about the way that cancer is treated and the business of cancer drug development.
[08:49] and something shocking emerged from that exercise, which is that finance plays a huge role, sometimes way too big a role in how drugs get developed.
[09:01] and so that's what i want to talk to you about today.
[09:03] the the rather strange conundrum that at the very same time that all of these incredible scientific breakthroughs are being made, there are nevertheless enormous
[09:13] bottlenecks in the system that could actually be addressed with better financing.
[09:19] So let me start by observing something uh important about where we are in the biomedical field and that is the fact that right now we happen to be at an inflection point.
[09:33] And I want to describe to you what that means, what it means to be at an inflection point in biomedicine through two examples.
[09:39] The first example has to do with these two kids, Caroline and Cole Carper.
[09:46] These kids were born with a relatively rare genetic disorder called Libers congenital amaurosis.
[09:53] It's a genetic typo that ultimately causes their retinas to degrade.
[10:01] Shortly after birth and so within a few months of birth, they've become blind.
[10:08] And the amazing story about Caroline and Cole is that they participated in a clinical
[10:14] trial a few years ago with a company called spark therapeutics.
[10:18] developed out of the university of pennsylvania in philadelphia.
[10:21] and in 2016 spark received fda approval for a gene therapy.
[10:28] a therapy that takes the correct version of the gene that was altered in caroline and cole's case.
[10:38] the correct version is inserted into a virus and that virus is injected into the back of the eyes of these two young kids and in a matter of days they could see again.
[10:50] it's really quite extraordinary.
[10:54] for example when when caroline was asked about what it was like she said i went outside when it was snowing and i was like oh i can see the snowflakes.
[11:05] it was really cool to actually see something that i've never seen in my life before.
[11:10] it's a one-time injection of this viral vector.
[11:16] into their eyes and now they can see.
[11:20] the second example I want to give you is a case study that I wrote about a company called Agilis Biotherapeutics.
[11:26] This is a company that was started up a few years ago in here in Cambridge, Massachusetts and it was a company that was designed to take advantage of some research that was done by a Taiwanese doctor by the name of Paul Who.
[11:43] Doctor Who studied a very rare condition known as L-amino acid decarboxylase deficiency.
[11:50] That's a mouthful, I realize it's a genetic typo that causes infants to not be able to produce dopamine, a pretty important neurotransmitter.
[12:03] I'm sure you've all heard of the the runner's high experience.
[12:07] I've never experienced it.
[12:08] Usually when I run, I just get exhausted.
[12:10] But people tell me that if you run long enough, you actually start feeling good.
[12:14] It's because dopamine begins to be released in your brain.
[12:18] children with aadc deficiency cannot produce dopamine.
[12:24] and this has some pretty significant effects.
[12:26] because in addition to providing you positive reinforcement in various decision-making contexts, dopamine is critical for developing motor functions.
[12:36] and so infants that are aadc deficient you don't actually know that there's any throng until after a few months when you realize that your baby is not lifting his head up, is not moving his arms, is not rolling over, is not sitting up, it's not doing all of the things that a normal baby would do.
[12:55] so doctor who realized that he could actually fix this genetic typo by doing the same kind of technique as we saw before with spark.
[13:02] take the correct version of the gene insert it into a virus, take the virus inject it into the brain and that virus will cause the gene to be replaced and therefore begin to start producing.
[13:19] dopamine.
[13:21] I'm going to show you a video clip of patient number four in the clinical trial for this particular therapeutic.
[13:29] And um I think you'll you'll you'll see it's quite striking.
[13:33] The the beginning of the video clip will show you the infant, uh, who is actually not an infant when he was diagnosed, he's two years old.
[13:41] And we'll see the baseline before the treatment.
[13:44] And then we'll see what happens one year after this one-time injection.
[13:47] So um, so here's the infant baseline.
[13:56] [Music]
[13:57] And you can see that the infant can't raise his head.
[14:00] Can't lift his arms without any kind of help.
[14:03] Can't roll over.
[14:05] He's basically uh confined to that kind of a unfortunate circumstance.
[14:11] One year later, this is the same child moving, crawling.
[14:14] Can't quite stand yet because obviously he
[14:20] just learned how to uh engage in some of these motor functions recently.
[14:25] so you can see that he's still somewhat unsteady.
[14:31] but now take a look at what happens a year later.
[14:36] so this is now two years after the treatment he's four years old.
[14:41] and he can stand he can't quite walk normally yet.
[14:45] and uh we're not sure whether he'll ever walk normally because it's important to engage in these motor functions you know at the earliest stages of life.
[14:54] and he only got the ability at two years old.
[15:00] but from his parents perspective this is an absolute miracle that he's able to do this.
[15:09] the blind shall see and the lame shall walk.
[15:12] you know that's a phrase out of religious text but it's happening today.
[15:21] thanks to these incredible therapeutics.
[15:24] now by the way this little biotech company.
[15:25] the reason that i got in touch with them was because they were struggling for funding and read some of my research and wanted to get some help.
[15:33] and uh so i wrote the case study for them.
[15:34] and uh it turns out that after the particular phase two trial where they showed this video clip.
[15:42] this little biotech company was acquired by a much larger one for about 900 million dollars.
[15:50] one therapy 10 patients 900 billion dollars.
[15:54] so this goes to marcus's point about how this ecosystem is incredibly important.
[15:59] the bigger fish will eat the smaller fish and in this case the smaller fish are perfectly happy to be eaten because it basically allows them to take these therapeutics and develop them and distribute them to a broader audience.
[16:12] so this is the backdrop for what i mean by biomedicine is at an inflection point.
[16:18] and a few years ago my mit colleagues phil sharp susan
[16:22] hockfield and tyler jacks published this report called convergence the convergence of the life sciences physical sciences and engineering.
[16:32] the point that all of these disparate fields are now coming together to create tremendous progress in the life sciences for medical interventions.
[16:42] and one way that experts in this field describe these inflection points is is this omics revolution.
[16:51] genomics the study of the sequence of the human genome.
[16:57] epigenomics the study of the on off switches that cause certain genes to be expressed and others to be suppressed.
[17:04] transcriptomics the study of how these gene sequences get translated into proteins.
[17:10] proteomics the study of the 20 to 25 000 different proteins that make up the human body.
[17:14] metabolomakes the study of all the chemical reactions that occur to make life possible.
[17:20] and most recently microbiome makes the study of the bacterial colonies that inhabit the body and provide us.
[17:23] with all sorts of important functions.
[17:26] all of these omics have experienced tremendous advances over the last few years with the exception of one and that exception is economics.
[17:37] the fact that we need to figure out better ways of paying for all of these therapeutics, that's where the bottleneck occurs.
[17:46] this is not to say that economists haven't made a lot of progress, we certainly have.
[17:50] but in terms of applying these ideas to the healthcare system, there's a lot that we can do and that's what i want to talk about today.
[17:58] so in particular i want to talk about what's called the valley of death.
[18:02] the idea that there is a certain set of bottlenecks, particularly in the early stages of drug discovery when a drug is being developed in a laboratory, it turns out that there's a lot of risk, a lot that we don't know.
[18:16] but as it gets put into humans in clinical trials phase one, phase two, phase three as it progresses, it becomes de-risked.
[18:23] another term that the
[18:25] biopharma industry uses.
[18:27] and by the time it reaches phase two or phase three there's a lot of money.
[18:31] that's around biopharma companies are able to fund those late stage trials fairly easily but the early stages of drug discovery is really the valley of death.
[18:42] that's what people call it in the industry.
[18:45] and the question that i wanted to ask as an economist is why why is there this valley of death.
[18:49] i just assumed that if there was some real medical needs and some smart biomedical experts that can figure out how to deal with it that the money will just magically show up.
[18:59] and the fact is that it's because of these incredible breakthroughs because of these breakthroughs that sometimes funding is not available and it has to do with increasing risk and uncertainty.
[19:11] you see it's a very counter-intuitive at least from my perspective.
[19:15] typically in finance when we learn more about an investment or a particular security the more we learn the smarter we get the lower the risk.
[19:27] that's not true in biomedicine.
[19:30] the smarter we get about biomedicine,
[19:31] the more risky in some cases are the investments that are associated with it.
[19:39] because as we learn about these new innovations it gives all sorts of new opportunities for young upstarts to make obsolete the existing technologies that are generating revenues for companies and their investors.
[19:53] and if there's one thing i can tell you as a financial economist it's that investors don't like risk.
[20:01] so i want to give you an example of that that i use with my first year mba students.
[20:06] nothing to do with with with biomedicine or healthcare.
[20:11] it's about risk and reward.
[20:13] and the experiment that i do with my mba students is to show them four financial investments.
[20:17] i don't tell them what they are or even over what time period they spend i just show them what happens if you put a dollar in each one of these investments the
[20:28] green one turns a dollar into two
[20:29] dollars
[20:30] but not particularly rewarding not
[20:32] particularly risky
[20:34] the red line turns a dollar into five
[20:36] more
[20:37] rewarding but more volatile the blue
[20:39] line
[20:40] is the most profitable but also the most
[20:42] volatile and the black line is somewhere
[20:44] in the middle
[20:45] and i ask my students to choose
[20:48] one and only one of these investments
[20:51] and so i'm going to ask you to do that
[20:52] now if you don't mind i'm going to take
[20:54] a poll
[20:54] and ask all of you to choose one of
[20:56] these four investments
[20:58] for your retirement fund for your kids
[21:01] college education
[21:02] fund maybe you're managing money for
[21:03] your parents or grandparents if they're
[21:05] older
[21:06] just just pick one of these that you
[21:08] feel is
[21:09] the best trade-off between risk and
[21:11] reward clearly
[21:12] there's no right or wrong answer it's
[21:14] just a matter of what your preferences
[21:16] are your risk preferences
[21:18] and so there's a really interesting
[21:20] phenomenon
[21:21] that all of us exhibit in thinking about
[21:23] risk and reward and this is the
[21:25] element that is pretty critical for
[21:28] understanding what's going on in the
[21:29] biomedical field
[21:31] so i'm going to just give it a few more
[21:33] seconds and
[21:34] i'm going to close the poll and i think
[21:37] you'll see
[21:38] an interesting pattern here that that
[21:40] i've seen
[21:41] in virtually all of my classes that i've
[21:43] taught
[21:44] there's actually only one one exception
[21:47] in terms of
[21:49] the particular choices that people make
[21:51] and i'll tell you about that
[21:52] in a few minutes okay so
[21:56] most of you have responded so i'm going
[21:58] to end the poll now
[21:59] and let me just share the results
[22:01] because
[22:02] what we see here is basically what we've
[22:04] seen
[22:05] in virtually every context that i've
[22:08] ever presented this
[22:09] problem and that is that the by far the
[22:12] most popular choice
[22:13] is the black line why because it seems
[22:16] like it's got the best trade-off between
[22:18] risk
[22:18] and reward right it's not the most
[22:19] rewarding but it's a lot less volatile
[22:23] the other choices
[22:25] okay so let me uh let me tell you now
[22:28] um what you all pick first of all the
[22:30] time period
[22:32] uh runs from 1990 to 2008. so we're
[22:36] talking about
[22:37] you know close to a 20-year period is a
[22:39] long time
[22:41] the green line which only three percent
[22:43] of you chose
[22:44] is u.s treasury bills the safest asset
[22:47] in the world
[22:48] at least for the next few weeks we'll
[22:50] see what happens with the budget
[22:51] discussions and uh
[22:52] hopefully we'll be able to continue on
[22:53] with this but not very rewarding if you
[22:56] put your money in t-bills
[22:58] as of 2008 you would have earned pretty
[23:00] much nothing
[23:02] since then now how about the red line
[23:05] some of you picked it it wasn't terribly
[23:06] popular but some of you picked the red
[23:08] line the red line
[23:09] is the u.s stock market the s p 500 and
[23:12] if you had picked that well
[23:14] congratulations you would have done just
[23:16] fine since
[23:17] 2008. the blue line which actually a
[23:21] surprising number of you did choose i'm
[23:23] actually quite impressed
[23:25] that is the single pharmaceutical
[23:27] company pfizer
[23:29] way more volatile but way more rewarding
[23:33] and if you had picked pfizer
[23:35] congratulations you would have done
[23:37] even better than than the rest
[23:40] now what about the most popular choice
[23:42] the black line
[23:44] well the black line is the returns to a
[23:47] private investment called the fairfield
[23:48] century fund
[23:50] that was the feeder fund for the bernie
[23:52] madoff ponzi scheme
[23:54] which is why i had to stop this
[23:55] experiment in 2008 that's when the ponzi
[23:57] scheme blew up
[23:59] now i i apologize i know that you're
[24:02] very frustrated um
[24:04] my students get annoyed as well the
[24:06] purpose of this exercise
[24:08] is to illustrate why it is that this
[24:11] ponzi scheme
[24:12] ultimately ended up suckering in so many
[24:15] people
[24:15] 50 billion dollars worth by the way and
[24:18] it's because
[24:19] it is human nature we are all attracted
[24:22] to investments
[24:24] that have high yield and low volatility
[24:28] and in finance we have a term for that
[24:30] it's called the sharp ratio the sharpe
[24:32] ratio
[24:33] is the excess return of an investment
[24:35] above
[24:36] and beyond t-bills per unit risk
[24:39] as measured by the standard deviation so
[24:42] if you calculate the sharp ratio for
[24:43] pfizer
[24:44] and for the s p 500 it's about the same
[24:46] you know 0.4.5
[24:48] and on paper the sharp ratio for the
[24:51] madoff ponzi scheme
[24:52] was an order of magnitude larger that's
[24:56] why so many of us were drawn to this
[24:59] scam it's because we're all looking for
[25:01] something
[25:02] that is high returning low volatility
[25:06] and sometimes things that are too good
[25:08] to be true aren't
[25:09] true it turns out
[25:12] that biomedicine has had a sharp ratio
[25:16] that has been going
[25:17] down over the course of the last decade
[25:21] down not because the returns are any
[25:23] less on the contrary the returns are
[25:25] great
[25:26] if you're successful the problem is that
[25:29] the complexity
[25:30] of biomedicine has grown and therefore
[25:33] it's much harder to tell now which is
[25:36] going to be
[25:37] a winner and which is going to be a
[25:38] loser
[25:40] so let me give you a second pop quiz the
[25:43] second pop quiz
[25:44] is a simple investment opportunity that
[25:46] i'd like you to just tell me whether you
[25:48] would or would not invest okay
[25:50] so i'm going to put up a poll in a
[25:51] minute but let me tell you what the
[25:52] characteristics are
[25:53] so the investment is 200 million dollars
[25:57] up front you need to give me 200 million
[25:59] dollars
[26:00] and there's a 10-year horizon before
[26:03] which you will see
[26:04] any kind of a return okay
[26:08] and one more thing the probability
[26:11] of any kind of a positive payout
[26:14] whatsoever
[26:15] is five percent with 95 probability
[26:20] you're going to get nothing back after
[26:22] 10 years
[26:24] so let me um start the poll
[26:28] and uh let's see how many of you would
[26:30] be willing to
[26:31] uh invest uh in this
[26:35] so um let me uh
[26:39] pull up poll number two and uh
[26:44] yeah so but then if it goes out well
[26:47] what's the return uh yeah that's a great
[26:49] question marcus i'm glad you asked i was
[26:51] hoping that somebody would ask this
[26:53] um and i will tell you in just a second
[26:55] let me just uh
[26:57] get this poll going i launched the wrong
[26:59] poll i apologize
[27:01] um let's see poll number two okay here
[27:04] we go
[27:05] so yeah let me tell you um it turns out
[27:10] oh for some reason it's still giving me
[27:12] poll number one
[27:16] let me end this poll and try
[27:20] it again apologies
[27:23] um share results
[27:29] and uh i can launch it yeah if you can
[27:32] launch poll number two that'd be great
[27:35] yeah great thank you so um
[27:38] usually most of my students
[27:42] don't even ask me the question what do i
[27:44] get
[27:45] if i succeed because they don't need to
[27:47] know
[27:48] all they need to know is that it's 200
[27:51] million dollars
[27:53] 10 years 95 failure rate
[27:56] and their answer is exactly what we're
[27:59] getting
[27:59] from the participants which is no thank
[28:03] you
[28:03] not interested now it turns out that
[28:07] what i'm showing you is the back of the
[28:09] envelope calculations for
[28:10] what it takes to develop a single
[28:12] anti-cancer compound
[28:14] and so if you actually do the math and
[28:18] you are successful
[28:19] in developing an anti-cancer compound
[28:21] and by the way the success rate
[28:23] uh is in fact only five percent
[28:26] historically it's a little bit lower
[28:27] than that
[28:28] if you are successful though you're
[28:30] going to get revenues
[28:31] on the order of about 2 billion a year
[28:35] from years 11 to 20. why is that
[28:39] it's because typically a biomedical
[28:41] patent lasts for 20 years
[28:43] if it takes 10 years to do clinical
[28:44] trials you've got another 10 years to be
[28:46] able to benefit from that
[28:48] and if you're successful you can
[28:50] generate 2 billion a year in years 11
[28:53] 12 13 all the way through 20 and then
[28:55] your your
[28:56] drug goes off patent so that amortizes
[29:01] to about 12.3 billion dollars
[29:05] in year 10 if you're successful
[29:08] and zero if you are not
[29:12] and so what that amounts to
[29:16] is a return of about 12
[29:20] on average the expected return during
[29:22] those first 10 years annualized
[29:24] and then a standard deviation of 423.5
[29:30] in other words a sharp ratio of about
[29:33] zero
[29:34] it's it's really really small because
[29:37] the risk is so high
[29:39] and so i'm going to end the poll now and
[29:41] we can see
[29:43] that um most of you are not interested
[29:48] and you're not interested because this
[29:50] is way too risky for the typical
[29:52] individual now i'm gratified that some
[29:55] of you
[29:56] said yes and i'm i'm i'm hopeful that
[29:59] we'll see more of that but it turns out
[30:01] that there's a an even more powerful way
[30:04] to be able to get to a medical question
[30:07] yes so with the vaccine everybody was
[30:09] expecting it took us about 10 years to
[30:11] develop a new vaccine as well but now we
[30:13] could do it in one year
[30:14] yeah you expect we can new technology
[30:16] will bring the 10-year horizon down
[30:18] to or absolutely no no not on the
[30:21] contrary
[30:22] it is absolutely happening so in the
[30:24] area of vaccines that's one example but
[30:26] there are many other examples in cancer
[30:28] in immunotherapy treatments for various
[30:30] different diseases in the gene therapies
[30:32] that i just showed you at the very
[30:33] beginning
[30:34] that's what i mean by the inflection
[30:36] point the economics are changing as the
[30:39] science is changing
[30:41] and the combination of having better
[30:43] science
[30:44] and better financing could actually make
[30:46] some tremendous
[30:47] tremendous impact on health care so it's
[30:50] exactly right
[30:52] so let me first then now that we
[30:54] acknowledge that the science is
[30:55] progressing
[30:56] let me tell you how finance can actually
[30:58] help
[31:00] in this case it turns out that instead
[31:02] of doing
[31:03] one project at a time imagine if we did
[31:07] 150 of them all at the same time
[31:10] now i realize that that's a tall order
[31:13] because we're going to need
[31:14] 150 times 200 million or 30 billion
[31:18] dollars of capital
[31:20] and where are we going to get 30 billion
[31:24] well um as an economist i have a very
[31:26] simple answer the answer is
[31:28] assume we have 30 billion dollars now
[31:31] i'll i'll come back to it tell you
[31:33] how we get that in a minute but let's
[31:34] suppose that we do it turns out that if
[31:36] we have 30 million dollars
[31:39] the economics change entirely
[31:43] imagine that the 150 projects are
[31:45] statistically
[31:46] independent i'll come back to that
[31:48] assumption a little bit later on but if
[31:49] they
[31:50] are independent and identically
[31:52] distributed or
[31:53] iid then the expected return stays the
[31:57] same of the portfolio
[31:59] but the risk goes down at the rate of
[32:02] the square root of n
[32:04] which means that the volatility is now
[32:07] 35 to give us a sharp ratio
[32:11] of 0.34 you see that's
[32:15] that's an incredible reduction
[32:18] in risk so andy can ask you if i go back
[32:22] to these two models of funding where i
[32:24] say
[32:24] traditionally it was this big companies
[32:27] doing many many
[32:28] projects and many vaccines while we
[32:30] moved actually to the startup model
[32:32] so this would be a strong argument to
[32:34] have these big companies
[32:36] you know having much lower funding costs
[32:38] and proceeding and
[32:39] developing 150 drugs simultaneously why
[32:42] did we see then this startup
[32:44] moved to the startup model is there some
[32:45] other force which makes the startup
[32:47] model way more
[32:48] successful there is absolutely that's a
[32:50] fantastic question
[32:52] the answer is the nature of innovation
[32:55] itself
[32:56] when you are a big pharma company it
[32:59] does not necessarily
[33:01] pay for you to take risks and innovate
[33:04] why because you have much higher sharp
[33:07] ratio activities to get involved in
[33:09] namely marketing distributing and
[33:12] licensing drugs
[33:14] drug discovery the early stage r d
[33:18] that is something that is best left for
[33:21] startups and so the question is is there
[33:24] a way
[33:25] that we can have our cake and eat it too
[33:28] can we figure out a way
[33:29] to finance startups and take that risk
[33:33] without having to become a big pharma
[33:35] company but at the same time still
[33:38] lowering the cost of capital and that's
[33:41] where finance comes in so
[33:44] can we really raise the 30 billion
[33:46] dollars now clearly pharma companies
[33:48] already have
[33:49] that kind of money if you look at pfizer
[33:51] gsk novartis
[33:52] they have enormous amounts of corporate
[33:55] cash
[33:56] but their incentives are different than
[33:59] smaller biotech companies and i'll come
[34:02] to that in a minute
[34:04] so can we really raise the 30 billion
[34:05] dollars well the answer is
[34:07] it depends it depends on the risk reward
[34:10] profile
[34:11] and so let me assume
[34:14] that we are statistically independent
[34:17] and let me assume
[34:18] that we have 150 projects and
[34:21] let me assume that we can actually
[34:24] finance
[34:25] a portion of this particular fund
[34:28] using bonds using debt financing
[34:32] now how much debt can i issue
[34:36] well the answer is it depends on what
[34:39] the yields are
[34:40] and what the risk profile is of your
[34:42] portfolio so let's talk about risk
[34:44] i've got 150 projects they're all
[34:47] independently and identically
[34:48] distributed with a five percent
[34:49] probability of success
[34:51] but if i'm successful in any one of them
[34:53] i make 12.3 billion dollars
[34:56] in year 10. so what's the probability
[35:00] that i have at least three successes
[35:03] out of 150 shots on goal to use a hockey
[35:07] or a soccer term
[35:08] well let's do the math we know that this
[35:11] is a binomial distribution
[35:13] because each one of these shots on goal
[35:15] is a bernoulli trial
[35:17] the probability of at least three hits
[35:20] out of 150
[35:21] independent tries each with a
[35:23] probability of success of five percent
[35:25] happens to be 98.18 percent
[35:31] and so if i issue
[35:34] debt of three times
[35:38] twelve point three billion dollars or
[35:41] thirty six point nine billion dollars if
[35:44] i issue
[35:46] debt with the face value of slightly
[35:49] less than
[35:50] 39 36.9 billion dollars what's my
[35:52] default rate
[35:54] the default rate is 1 minus
[35:58] the probability that i'll have at least
[35:59] three hits and that is
[36:01] a little less than two percent a default
[36:04] rate of a little less than two percent
[36:06] means that i've got a single a rating
[36:08] and as of a few days ago the single a
[36:10] yield
[36:11] was given by one point six four percent
[36:13] according to the bank of america merrill
[36:15] lynch index
[36:16] and at that yield i can actually raise
[36:19] more than 30 billion dollars to finance
[36:21] this portfolio
[36:23] okay marcus can you launch the third
[36:27] poll
[36:28] the question i want to ask is now how
[36:30] many of you
[36:31] would be willing to invest in this
[36:34] particular opportunity
[36:37] if i've got 150 projects all
[36:40] independently identically distributed
[36:43] five percent chance of success
[36:46] what is the likelihood that you will be
[36:48] willing to invest
[36:50] in this
[36:53] i have troubles launching the poll
[36:56] yeah let me uh let me see if i can do it
[36:59] from here
[37:01] uh poll number one
[37:04] two three uh
[37:08] and launch poll okay
[37:13] so how many of you will be willing to
[37:15] invest in this
[37:17] and i think you're gonna see a pretty
[37:19] interesting difference
[37:21] from poll number two
[37:24] the fact is that using financial
[37:26] engineering we have been able to change
[37:28] the risk reward trade-off
[37:30] and that has dramatic effects on the
[37:33] supply of funds
[37:36] so not surprisingly right now i'm
[37:38] looking at the results
[37:40] and about 92 percent of you polled
[37:43] have said yes you'd be willing to invest
[37:45] in this whereas
[37:47] the one-off the one-off investment less
[37:50] than
[37:51] a quarter of you would be willing to
[37:52] take that investment
[37:55] and with all of the other tricks
[37:59] of the financial trade things like
[38:02] securitization credit default swaps
[38:05] government guarantees we can do even
[38:08] better
[38:10] now you might uh so these are the
[38:13] results
[38:13] not surprisingly and uh i think you
[38:17] would you you
[38:18] understand now what the power of
[38:20] financial engineering
[38:21] is now you might be reacting to this by
[38:24] saying wait wait wait a minute
[38:25] isn't this what got us into all the
[38:27] trouble with the financial crisis of
[38:28] 2008
[38:30] and the answer is yes exactly
[38:34] these techniques played a prominent role
[38:36] in the financial crisis
[38:38] not because they don't work
[38:41] but because they work way too well these
[38:44] are incredibly powerful financial tools
[38:47] and if you apply them to focused
[38:51] objectives you can make enormous
[38:53] progress
[38:55] but you have to be careful and you know
[38:57] the fact that interest rates are at an
[38:59] all-time low this is a single a rated
[39:01] yield over the course of the last
[39:03] several decades it means that there's
[39:05] tremendous resources
[39:07] waiting to be tapped but you need to be
[39:10] able to use that
[39:11] energy in a controlled and responsible
[39:14] way
[39:15] now they can ask you a question so i
[39:17] would like to know
[39:18] how i made a comment essentially there's
[39:21] of course market risk
[39:22] and then there's idiosyncratic risk and
[39:24] this medical risk seems to be totally
[39:26] idiosyncratic
[39:28] can you elaborate on that a little bit
[39:30] do you see something
[39:31] that the big you cannot carry it through
[39:33] to the end because of some market
[39:35] disruptions is this the only market risk
[39:37] or what is uh
[39:38] i love that question can you hold on for
[39:40] just about
[39:41] five minutes because i'm going to get
[39:42] directly to that but yes that's an
[39:44] excellent question
[39:45] and it's one that i struggled with so
[39:47] we're going to get right to that
[39:48] but before i do i want to make the point
[39:50] about diversification which is related
[39:52] to that question
[39:53] and that is the assumption that i made
[39:56] that 150 projects are
[39:58] independently and identically
[40:00] distributed that is a
[40:01] tall order right that's a very very
[40:03] restrictive assumption
[40:04] and we know from the financial crisis
[40:06] that all of these mortgage-backed
[40:08] securities
[40:09] they failed the subprime market failed
[40:11] because
[40:12] these defaults were highly correlated so
[40:15] correlation matters a lot and i can show
[40:17] you
[40:18] why in the case of no correlation
[40:21] the probability of at least three
[40:23] successes
[40:24] out of a hundred and fifty tries each of
[40:27] them with a five percent probability of
[40:28] success
[40:29] is what i just described to you 99 98.18
[40:34] that's what this graph shows but now
[40:37] what happens if there's a
[40:38] 10 pairwise correlation
[40:42] in terms of the success and failure of
[40:44] these
[40:45] compounds well then the probability of
[40:48] at least three successes
[40:50] goes down to about 89
[40:53] if the correlation between successes is
[40:57] 40 now the probability of at least three
[41:00] successes
[41:01] is down to about 55 percent and if the
[41:04] pairwise correlation
[41:06] is 80 then you're now at a 23
[41:10] probability of at least three hits and
[41:13] at that
[41:13] level nobody is going to want to buy
[41:15] your debt you cannot finance
[41:18] by debt so the key is correlation
[41:22] and diversification when i described
[41:25] this to one of my pharma
[41:26] colleagues one of my colleagues is in
[41:28] the pharma industry
[41:30] he said you know what this is exactly
[41:31] what i've been telling my colleagues
[41:33] but the way that i put it is using third
[41:36] grade soccer as the analogy
[41:39] and i i said what do you mean i i don't
[41:41] play soccer my kids didn't play soccer
[41:43] and he said well if you've ever coached
[41:45] third grade soccer which is what i do
[41:46] for my daughter's team
[41:48] this is what third grade soccer looks
[41:50] like all the kids crowding around the
[41:52] same ball
[41:53] and what i tell them is no no spread out
[41:55] spread out go to where the ball
[41:57] will be not where the ball is and when
[42:00] you compare
[42:01] third grade soccer to world cup soccer
[42:04] you see the difference the fact is
[42:07] that the biopharma industry does tend to
[42:10] crowd
[42:11] around the same set of ideas there's a
[42:14] reason for that they're not dumb
[42:15] obviously
[42:16] the reason that they do so is because
[42:17] they want to make use of the scientific
[42:19] knowledge that they have
[42:20] and there's benefits to being able to
[42:22] work together collaboratively on the
[42:24] same set of ideas
[42:26] but what makes sense from a scientific
[42:28] point of view does not necessarily
[42:30] make sense from an investor's point of
[42:32] view and so diversification
[42:34] is key if we can better diversify we can
[42:37] lower the cost of capital
[42:39] and now i'm going to get to the question
[42:41] that marcus just asked
[42:43] about diversification and the cost of
[42:45] capital
[42:46] so and this goes to the last poll that i
[42:48] wanted to
[42:49] issue and that is we know
[42:53] from models like the capital asset
[42:55] pricing model
[42:56] that the cost of capital the required
[42:58] rate of return
[43:00] is directly related to how much risk
[43:02] you're bearing
[43:03] but as the person in the chat asked
[43:05] about systematic or idiosyncratic risk
[43:08] it matters a lot as to what kind of risk
[43:10] we're talking about
[43:12] because what matters for the cost of
[43:14] capital
[43:15] is systematic risk it's beta it's the
[43:18] correlation
[43:19] with the market portfolio right so the
[43:22] higher
[43:22] the correlation the higher the cost of
[43:25] capital
[43:26] and you would think with biomedicine the
[43:29] correlation
[43:30] of a drug to the s p 500 should be
[43:34] virtually zero right because it's it's
[43:36] scientific risk
[43:38] so i want to ask you one last question
[43:42] which is to tell me what you think the
[43:45] systematic component of risk is
[43:48] in the biopharma industry and for the
[43:51] sake of argument
[43:52] i'm going to uh one as the benchmark for
[43:55] beta right so if you've got a beta of
[43:57] one
[43:57] you are as risky as the s p 500 and
[44:00] therefore you should carry
[44:02] a cost of capital or risk premium
[44:03] comparable to the market portfolio
[44:06] historically the risk premium has been
[44:07] about eight point three percent
[44:09] most people think going forward it
[44:11] should be closer to uh six percent
[44:13] cost of capital if you add a treasury
[44:15] bill yield of you know two percent for
[44:17] ten-year yield
[44:18] you're getting an eight percent cost of
[44:19] capital right
[44:21] so the poll that i want to ask you is to
[44:23] just to tell me what do you think the
[44:25] beta is
[44:26] of the pharma industry which is focused
[44:28] on late stage
[44:30] drug development and marketing
[44:32] distributing and licensing or the early
[44:34] stage
[44:35] biotech industry the startups that are
[44:37] focusing on the really innovative
[44:39] the really risky kind of science and
[44:42] medicine
[44:43] so let me launch uh this poll and let's
[44:46] let's get your thoughts and i'll tell
[44:47] you
[44:48] uh what my thinking is uh you know once
[44:50] we uh once we do that
[44:52] oh sorry i didn't share the results from
[44:53] before i apologize um
[44:55] but i think you knew what the the
[44:57] results would be um
[44:59] okay so let me stop that and then launch
[45:01] the last poll
[45:04] so the last poll is a two-part question
[45:07] uh it's basically asking you to pick
[45:10] what the beta is
[45:11] of the pharma industry in the first part
[45:13] of the question
[45:14] and then tell me what do you think the
[45:16] beta is of the biotech
[45:18] industry in the second part of the
[45:20] question because
[45:21] the biotech industry is really where
[45:22] that kind of scientific risk
[45:24] is the greatest the pharma industry
[45:27] is really focusing on the marketing
[45:31] distributing and licensing of drugs
[45:32] that's much more late stage and the
[45:34] sharp ratio
[45:35] actually tends to be you know pretty
[45:36] high the question is
[45:38] what about the biotech industry what do
[45:41] you see
[45:42] as the risks and therefore the cost of
[45:44] capital
[45:45] in the biotech industry after i get your
[45:47] answer i'm going to show you the results
[45:48] and i think you'll be you'll be
[45:50] surprised at least i was i was
[45:52] really surprised by the results now just
[45:55] to let you know as a backdrop most
[45:56] people think that healthcare
[45:58] is a really great diversifier right it's
[46:00] a really great investment because
[46:02] you know the risks don't seem to be
[46:04] correlated with anything that
[46:06] we would care about in the world of
[46:08] finance
[46:10] so let me stop the poll here
[46:14] and share the results and i think you'll
[46:17] see
[46:18] what you know we were talking about so
[46:21] for the pharma industry um the majority
[46:25] of you
[46:26] felt that the beta was somewhere between
[46:28] 0.5 and 1
[46:29] and between 1 and 1.5 those are the two
[46:32] most popular
[46:32] responses the biotech industry
[46:36] most of you felt that the beta was
[46:38] smaller right the majority of you said
[46:40] that the beta is between minus 0.5
[46:42] and plus 0.5 so let me uh
[46:46] stop sharing the results and tell you
[46:47] what the answer is so
[46:49] regard to the pharma industry you
[46:53] are absolutely correct that when you
[46:55] look at the beta of the pharma industry
[46:58] for the most part it's been below one
[47:01] and the reason is that the the
[47:04] the price elasticity of demand is
[47:07] generally pretty low for drugs right you
[47:09] need your medication regardless of
[47:10] whether the economy is in a boom or a
[47:12] bust
[47:13] and so pharma companies generally have
[47:16] had a pretty reliable
[47:17] uh risk premium and rate of return and
[47:20] therefore the cost of capital for pharma
[47:22] is on the order of eight to ten percent
[47:26] now let me show you the shocking result
[47:30] this is the beta for the biotech
[47:32] industry
[47:33] and while now it is slightly below one
[47:37] in fact historically when you look at
[47:39] all the biotech companies
[47:40] in the entire industry it is way
[47:44] higher than one on average the typical
[47:46] biotech company has a beta
[47:47] of about one and a half to two
[47:51] so i was shocked by this result
[47:54] because it's exactly as the person who
[47:57] asked the question
[47:58] said the risk of these biomedical
[48:01] experiments
[48:02] have nothing to do with the s p going up
[48:05] or down
[48:06] why should they have a higher beta and
[48:08] it took me a while to figure out the
[48:10] answer
[48:10] and the answer once i figured it out so
[48:12] obvious i felt quite embarrassed
[48:15] the answer is that a biotech company
[48:17] faces
[48:18] two kinds of risk the first kind of risk
[48:21] is scientific
[48:22] and the scientific risk is indeed
[48:24] totally uncorrelated
[48:26] with financial considerations however
[48:30] the second kind of risk is financial
[48:33] risk
[48:34] because biotech companies have to pay
[48:36] for stuff
[48:38] and so if you are in the midst of a very
[48:40] important experiment
[48:42] and you run out of money well guess what
[48:45] that experiment is not going to get done
[48:48] and so
[48:49] what we're seeing in the biotech
[48:50] industry is that
[48:52] financing risk often trumps
[48:56] scientific risk in terms of its cost of
[48:58] capital
[48:59] and effectively what biotech risk is is
[49:02] rollover
[49:03] risk the fact that at some point
[49:06] if you run out of cash and you need to
[49:08] refinance
[49:09] if markets are in a recession and you
[49:12] can't refinance
[49:13] you're out of luck and unlike other
[49:16] kinds of
[49:17] industries and assets a biomedical asset
[49:20] is really
[49:21] really driven by the underlying people
[49:23] that are involved
[49:25] and so if you've got expertise to
[49:27] develop a drug
[49:28] and you run into a cash flow problem and
[49:30] people start to leave
[49:32] the value of that asset gets destroyed
[49:35] unlike
[49:36] a parking lot if you buy a parking lot
[49:38] and you go bankrupt and you can't afford
[49:40] to develop it
[49:40] somebody else will take the parking lot
[49:42] and develop it there's very little loss
[49:44] in value
[49:45] tremendous loss and value when it comes
[49:47] to complicated intellectual property in
[49:50] the life sciences
[49:51] but why is it the structure this way why
[49:53] can't you issue more long-term
[49:55] debt or equity instruments and have to
[49:57] rely on rollover
[49:58] that's that's exactly what i'm
[50:01] suggesting yes
[50:02] so the fact is that the way that biotech
[50:05] has been financed historically
[50:07] is by venture capital venture capital is
[50:10] a phenomenal business model
[50:12] for financing an app a piece of software
[50:16] something where the proof of concept is
[50:19] a few hundred thousand
[50:20] to a couple of million and it'll take 18
[50:23] to 24 months
[50:24] that's why silicon valley has become
[50:26] such a dominant force
[50:27] they've got tremendous ability to
[50:30] finance these
[50:31] companies a biotech company
[50:34] requires tens of millions of dollars
[50:36] before you get to a proof of concept
[50:38] it requires 5 to 10 years not 24 months
[50:42] and the failure rate particularly in
[50:45] oncology
[50:46] is much much higher than in social media
[50:49] or software and so the business model of
[50:53] venture capital
[50:54] is the wrong model for financing
[50:56] biomedical innovation
[50:58] we need a different business model we
[51:00] need longer term financing
[51:02] financing that matches the horizon of
[51:05] the underlying asset and that's where
[51:07] securitization
[51:08] and and permanent capital and using
[51:11] these kinds of holding company
[51:13] structures come into play
[51:16] now i'm sure that there are many other
[51:17] questions that are in the chat window
[51:19] do we really need 30 billion dollars
[51:21] what's the market failure why hasn't
[51:23] this been done already
[51:24] isn't pharma already doing it if not
[51:25] shouldn't the government do it so on and
[51:27] so forth
[51:28] and the short answer to all of these
[51:31] frequently asked questions
[51:33] is i have no idea i'm just a financial
[51:36] economist what do i know about these
[51:38] things
[51:39] but i started asking people who should
[51:41] know
[51:42] i asked for example you know pharma
[51:44] executives
[51:45] and their view is that well we don't
[51:47] know we use these financing methods
[51:49] because that's what's been done in the
[51:50] industry that's how the biotech industry
[51:52] developed it was really through venture
[51:53] capital
[51:54] so then i started asking scientists and
[51:56] clinicians and their view is i have no
[51:58] idea
[51:59] i i just do the science and the medicine
[52:01] i'm not interested in the finance
[52:04] and the long answer to the question
[52:07] about all of these different issues is
[52:10] what i've been working on with
[52:11] collaborators over the last 10 years
[52:13] a variety of papers that i've written
[52:14] you can check out my website if you're
[52:16] interested in them
[52:17] but the long answer is that
[52:20] i think it is possible to do this and
[52:23] there are a number of ways
[52:24] that it's being done now so you know in
[52:27] the last few minutes
[52:28] let me start wrapping up by telling you
[52:31] about how this
[52:32] actually is being done so first of all
[52:34] in terms of the amount of capital that
[52:36] we need
[52:36] the answer of 30 billion really depends
[52:39] upon the particular projects
[52:41] you're working on the cost per per shot
[52:44] the probability of success duration
[52:46] of the trials so on and so forth all of
[52:48] that is what
[52:49] leads to how much money you need and in
[52:52] a paper that i recently
[52:53] posted on an ssrn
[52:56] with source code that allows you to run
[52:59] the simulations yourself
[53:00] you can actually determine how much
[53:02] capital you need depending on the
[53:03] particular medical area that you're
[53:05] focusing on
[53:06] but the key issue to all of this is that
[53:08] that finance
[53:10] experts and biomedical experts need to
[53:12] collaborate because
[53:14] no one group knows the answer to all of
[53:17] these questions
[53:18] you need to work together to figure out
[53:20] all of these different parameters to
[53:21] plug them into the software
[53:23] and that's really where i think the the
[53:26] value can be created
[53:27] by building bridges between these two
[53:29] fields
[53:30] i'll give you one example and then wrap
[53:32] up the one example
[53:34] where this method has actually been used
[53:37] has to do with orphan
[53:38] diseases rare diseases kind of like the
[53:41] diseases that i scribed at the very
[53:42] beginning of the talk
[53:44] things like hemophilia cystic fibrosis
[53:46] als gaucher syndrome
[53:48] all of these are examples of a rare
[53:51] disease which
[53:52] according to the 1983 orphan drug act in
[53:54] the united states
[53:56] are diseases that affect two hundred
[53:57] thousand patients or less
[54:00] so it turns out that because of these
[54:03] diseases
[54:03] uh the the each of them are relatively
[54:06] small but the fact that there are so
[54:07] many of them
[54:08] over 30 million americans suffering from
[54:10] over 7 000 rare diseases
[54:12] this is actually a really important area
[54:15] and
[54:16] due to a variety of policy as well as
[54:20] regulatory and economic and scientific
[54:22] issues
[54:23] it turns out that you don't need 30
[54:24] billion dollars to deal with this
[54:27] you need maybe four to five hundred
[54:29] million dollars you don't need 150
[54:31] projects you can actually get
[54:32] diversification with only
[54:34] 10 or 20. and why is that
[54:37] it's because the assumption of lack of
[54:41] correlation that i made
[54:42] in the previous example that actually
[54:46] holds pretty closely in the case of rare
[54:48] diseases
[54:50] and it does so because these rare
[54:52] diseases are so different
[54:54] and the therapeutics are so different in
[54:57] most cases not in all cases
[54:59] but in most cases they're so different
[55:01] that the success or failure of one of
[55:02] them
[55:03] generally has nothing to do with another
[55:06] one of them
[55:07] and so the pairwise uncorrelated
[55:09] assumption that i made earlier on
[55:12] that actually applies most closely in
[55:14] the case of orphan diseases
[55:17] and so when i published a paper in 2015
[55:20] and ran a simulation of a rare disease
[55:22] portfolio of 20 assets
[55:25] that cost about five or six hundred
[55:26] million dollars we were getting rates of
[55:29] return
[55:29] of on the order of 22 for the equity
[55:33] tranche
[55:34] and the comparable returns
[55:37] for the appropriate risk adjusted
[55:38] ratings for the
[55:40] debt tranches so this illustrates that
[55:44] in the case of rare diseases
[55:45] you can actually do really well with a
[55:48] relatively small
[55:49] portfolio and it's really because of
[55:51] that lack of correlation
[55:53] now you know if you want to know how
[55:55] unusual that lack of correlation is
[55:58] ask yourself the following question can
[56:00] you name
[56:02] five financial investments that are
[56:04] mutually
[56:05] uncorrelated i don't believe you can
[56:09] you could probably name two maybe three
[56:12] but you can't name five i can construct
[56:15] a portfolio of
[56:17] 20 very attractive rare disease
[56:20] therapeutics that are pairwise
[56:22] uncorrelated and if that's the case that
[56:24] means that you can increase the sharpe
[56:26] ratio of this portfolio
[56:28] at the rate of the square root of n
[56:30] where n is the number of projects
[56:33] now when i published this paper i was
[56:36] called by one of my former students a
[56:38] fellow by the name of neil kumar
[56:40] who was a chemical engineering student
[56:42] at mit
[56:43] and read the paper and said you know gee
[56:45] this is really interesting
[56:47] um would you mind re-running your
[56:49] simulations
[56:50] but not with your assumptions which you
[56:52] got from the industry literature
[56:54] use my assumptions instead so i said to
[56:57] neil
[56:58] sure i'd be happy to do that and
[57:01] over the course of four or five months
[57:04] we ran
[57:04] more and more simulations each time the
[57:06] results kept getting better and better
[57:08] because he was using you know his
[57:10] industry knowledge
[57:11] to make the appropriate assumptions
[57:13] about probability success and so on
[57:15] and what we do at the end of that
[57:19] process what we
[57:20] concluded was that this was actually
[57:22] feasible so he came into my office one
[57:24] day and said
[57:25] i just want to let you know i quit my
[57:27] job yesterday
[57:28] i'm going to do this and i have to tell
[57:30] you that that that scared the hell out
[57:32] of me i've never had that effect on any
[57:34] of my students
[57:34] i said to neil you sure you'd want me to
[57:36] run a few more simulations before you
[57:38] make that jump you had a small child and
[57:40] so i was really nervous for him
[57:42] he said no no no i'm really excited
[57:43] about this and so he started a company
[57:46] called bridge bio
[57:48] and i was a small seed investor in the
[57:50] company
[57:51] i wish i had invested a lot more now in
[57:53] retrospect
[57:55] but very shortly after the friends and
[57:57] family
[57:58] kind of funding we were able to raise
[58:00] money from
[58:01] a venture capitalist actually a private
[58:04] equity investor kkr
[58:06] that usually doesn't invest in these
[58:08] kinds of
[58:09] of companies because they don't invest
[58:12] in other people's funds
[58:13] because they are a fund but but bridge
[58:16] bio
[58:17] was not a fund it's an operating company
[58:20] that had subsidiaries that developed
[58:22] various different therapeutics so it was
[58:24] a portfolio
[58:25] a different business model a portfolio
[58:27] but with a company structure that
[58:29] allowed the
[58:29] underlying individual companies to
[58:31] innovate and
[58:33] over the course of the next few years we
[58:35] were able to raise
[58:36] more funding funding this is neil here
[58:38] neil kumar
[58:39] um over the course of about four years
[58:43] the company was able to raise about 700
[58:45] million dollars
[58:46] uh did an ipo in 2019
[58:50] the largest ipo at the time and um
[58:54] uh just last week announced yet another
[58:56] fundraising in addition to the ipo
[58:58] raising convertible bonds so now debt
[59:00] financing
[59:01] uh on the order of about 650 million
[59:03] dollars
[59:04] and as of a couple of days ago the
[59:07] company's market cap was about seven and
[59:09] a half billion dollars
[59:11] five years seven and a half billion
[59:14] dollar market cap
[59:15] but you know this is not what the
[59:18] company is most proud of
[59:20] the company is most proud of the fact
[59:22] that uh
[59:23] they put together a portfolio of about
[59:25] 20 projects
[59:27] of which four of them are now in phase
[59:30] three trials and we expect
[59:32] one of these to be approved by the end
[59:35] of this year and another one
[59:37] next year so therapies that would not
[59:40] have been developed
[59:41] except for the kind of funding that was
[59:44] uh
[59:44] raised by this company um
[59:48] and uh there are now more and more
[59:50] companies that are using these new
[59:51] business models
[59:52] increasing the scale because that's
[59:54] what's needed here it's the right scale
[59:56] and we don't yet have any securitized
[59:59] debt for biomedical portfolios but
[01:00:01] i'm hoping that sometime in 2021 we're
[01:00:04] going to see the first
[01:00:05] biobonds issued so let me wrap
[01:00:08] up by by saying that you know all of
[01:00:11] this all of these ideas and all of these
[01:00:13] methodologies
[01:00:14] was made very personal to me a few years
[01:00:17] ago when i ran into
[01:00:18] a colleague of mine in the biology
[01:00:20] department by the name of harvey lodesh
[01:00:23] when i heard about harvey's story i
[01:00:25] decided right then and there that i want
[01:00:27] to be
[01:00:28] harvey lodish and let me tell you why
[01:00:31] so in 1983 harvey was an assistant
[01:00:33] professor of biology at mit
[01:00:35] and he was asked by a venture capitalist
[01:00:37] to work with him to develop a treatment
[01:00:39] for a rare condition called gaucher
[01:00:40] syndrome that's also a single gene
[01:00:43] mutation
[01:00:44] a typo that prevents your body from
[01:00:45] producing some important enzymes
[01:00:48] and without those enzymes fatty acids
[01:00:49] build up in your bone marrow
[01:00:51] in your spleen and other organs and by
[01:00:53] the time you become a teenager
[01:00:56] you're dead it's a it's a very
[01:00:58] debilitating
[01:00:59] and deadly disease for a portion of
[01:01:01] gaucher patients
[01:01:03] and so harvey developed a method for
[01:01:05] replacing
[01:01:06] the enzyme that's missing in your body
[01:01:08] and in 1991
[01:01:10] the drug ceredaze was approved and
[01:01:13] since that time that drug and various
[01:01:16] improvements have saved the lives
[01:01:19] of at this point hundreds of thousands
[01:01:20] of patients around the world
[01:01:24] the company ultimately did very well it
[01:01:26] was bought out
[01:01:27] uh in i think 2010 uh for about 20
[01:01:31] billion dollars you may have heard of
[01:01:32] that company that harvey helped to start
[01:01:34] it's called genzyme
[01:01:36] and uh cenophy bought them for about 20
[01:01:39] billion dollars
[01:01:41] but that's not why i want to be harvey
[01:01:43] lodish although that's not a bad reason
[01:01:46] i want to be harvey because of what
[01:01:47] happened in 2002.
[01:01:50] in that year harvey's daughter was
[01:01:52] pregnant with her second child
[01:01:54] uh harvey and his wife's second grandson
[01:01:57] a boy named andrew
[01:01:58] great name by the way andrew was
[01:02:01] diagnosed
[01:02:02] in utero with gaucher syndrome
[01:02:08] what are the chances of that i asked
[01:02:11] harvey about this
[01:02:12] it was a very emotional conversation i
[01:02:14] asked him harvey in 1983
[01:02:17] when you started down this path to
[01:02:19] develop a drug for gauchers
[01:02:22] did you i have any idea that you carry
[01:02:24] the gene that that your grandson
[01:02:26] who is as yet unborn was going to have
[01:02:29] the disease did you have any idea
[01:02:32] and he said no i had no idea i i was i
[01:02:35] thought i was doing some cool science
[01:02:37] that could actually help some patients
[01:02:38] so that was you know all good
[01:02:42] it turns out that andrew in 2012
[01:02:45] developed the symptoms of full-blown
[01:02:47] gauchers
[01:02:48] but he's doing just fine thank you very
[01:02:50] much leading a totally normal life
[01:02:52] thanks to the drug that grandpa helped
[01:02:55] to develop
[01:02:56] this is why i want to be harvey lodesh i
[01:03:00] am not a md or phd in biology i don't
[01:03:03] have the ability to develop a drug that
[01:03:05] will help save
[01:03:06] my as yet unborn grandchildren but i
[01:03:09] realized something
[01:03:10] i realized that i could be harvey lodish
[01:03:13] we can all be harvey lotus
[01:03:15] if we invest in the therapeutics that
[01:03:18] will help
[01:03:19] save our as yet unborn grandchildren
[01:03:23] finance doesn't have to be a zero-sum
[01:03:25] game always
[01:03:26] at the right scale with the right kind
[01:03:29] of financing
[01:03:30] we can actually do well by doing good uh
[01:03:34] sorry i teach mba students so i have to
[01:03:36] make this a little bit more dynamic
[01:03:38] and so with all of your your interest
[01:03:41] and help
[01:03:42] in bridging the gap between finance and
[01:03:45] medicine
[01:03:46] i believe that we're going to be seeing
[01:03:47] some amazing therapeutics over the last
[01:03:49] few years we can all be harvey loadish
[01:03:51] and with that i want to thank you and uh
[01:03:53] happy to answer questions or
[01:03:54] uh discussion thanks a lot uh andy so we
[01:03:56] have a huge number of questions so i
[01:03:58] have to be selective apologize for the
[01:04:01] audience
[01:04:02] but extremely interesting and inspiring
[01:04:04] talk
[01:04:06] i wanted to i got the impression there
[01:04:09] is
[01:04:10] a private solution essentially it's more
[01:04:13] teaching
[01:04:14] educating developing the methods
[01:04:17] rather than ngos or the ppp
[01:04:20] partnership with the public sector is
[01:04:23] this the correct impression i got or is
[01:04:25] there any role for the government to
[01:04:27] intervene and help out or coordinate or
[01:04:31] can you elaborate on this space a little
[01:04:33] bit
[01:04:35] great that that's a great question and
[01:04:36] and there's absolutely
[01:04:38] a role for government in fact the reason
[01:04:41] that we have had such success in the
[01:04:43] private sector today
[01:04:45] is precisely because of government
[01:04:47] policy over the course of the last
[01:04:49] several decades i'll give you an example
[01:04:51] right now cancer has tremendous number
[01:04:54] of new drugs coming onto market
[01:04:56] i think last year there may have been
[01:04:58] like 27
[01:04:59] cancer drugs approved in one year one
[01:05:01] year so
[01:05:02] right now it looks like there's all this
[01:05:04] low-hanging fruit particularly
[01:05:06] applying gene therapy cellular therapy
[01:05:08] immunotherapy
[01:05:09] all of these new tools how did that
[01:05:12] happen
[01:05:14] it happened because in 1971 president
[01:05:17] nixon declared war on cancer
[01:05:19] and created all sorts of cancer centers
[01:05:22] around the country to fund
[01:05:24] basic scientific research in cancer
[01:05:26] biology
[01:05:28] which was not something that the private
[01:05:29] sector was going to do
[01:05:31] because there's no way to earn a rate of
[01:05:34] return because many of these ideas
[01:05:35] is just moving forward knowledge it's
[01:05:37] not necessarily patentable
[01:05:39] and so a lot of these ideas ultimately
[01:05:42] fed into the low-hanging fruit that we
[01:05:44] are benefiting from today
[01:05:46] there's estimates that the nih has over
[01:05:49] the course of the last
[01:05:51] 40 years put about a hundred billion
[01:05:53] dollars
[01:05:54] of money out there to develop cancer
[01:05:56] biology
[01:05:57] so we don't get these wonderful pieces
[01:06:00] of low-hanging fruit
[01:06:01] unless we plant the seeds decades before
[01:06:04] that's where government policy plays an
[01:06:06] important role
[01:06:08] another example is certain areas of not
[01:06:10] necessarily market failures
[01:06:12] but market challenges and i'll give you
[01:06:15] an example
[01:06:16] alzheimer's alzheimer's is clearly a
[01:06:19] really
[01:06:19] important problem over 5 million
[01:06:22] americans suffer from alzheimer's and
[01:06:24] last year
[01:06:25] we spent just dealing with the medical
[01:06:27] needs of alzheimer's patients
[01:06:29] medicare and medicaid spent 200 billion
[01:06:32] dollars
[01:06:33] on alzheimer's patients alone okay so
[01:06:35] this is a serious problem
[01:06:37] what's the probability of success for
[01:06:39] developing an alzheimer's drug well you
[01:06:41] can estimate it
[01:06:42] and if you use the last 10 years of data
[01:06:44] the answer
[01:06:45] is the probability of success is zero
[01:06:48] the last time we had an alzheimer's drug
[01:06:51] developed
[01:06:52] successfully approved by the fda was
[01:06:55] 2003.
[01:06:56] so it's been 18 years almost two decades
[01:07:00] before we have a single new alzheimer's
[01:07:02] drug despite the fact that we've got
[01:07:04] five million patients
[01:07:05] why it's because we don't understand the
[01:07:07] biology of alzheimer's
[01:07:09] and that's an example where government
[01:07:11] policy can play a huge role
[01:07:13] and it has so we declared a kind of a
[01:07:16] war
[01:07:17] on alzheimer's with the brain initiative
[01:07:19] that would it was launched by
[01:07:20] barack obama a few years ago and
[01:07:24] now the funding for alzheimer's at nih
[01:07:27] has gone up by several fold so
[01:07:30] i expect that in another five to ten
[01:07:32] years we're going to see intellectual
[01:07:34] property being generated that will then
[01:07:36] turn into valuable companies
[01:07:38] and there may be an alzheimer's drug
[01:07:41] approved
[01:07:41] sooner rather than later because of that
[01:07:43] effort so that's an example
[01:07:45] of government policy government policy
[01:07:48] can actually
[01:07:49] lower the risk and raise the sharp ratio
[01:07:52] of these investments by being able to
[01:07:54] provide funding
[01:07:56] for things that will create the kind of
[01:07:58] understanding the knowledge the
[01:08:00] intellectual property
[01:08:01] that the private sector can then take
[01:08:03] the rest of the way
[01:08:06] so you focused very much on the u.s how
[01:08:08] it's structured in the u.s if you take a
[01:08:10] more international perspective do other
[01:08:12] countries have a different way to so i
[01:08:14] said oh the early phase should be done
[01:08:15] by the government
[01:08:16] and then the later phase should be done
[01:08:18] by the private sector
[01:08:20] ideally some people might argue you know
[01:08:23] that
[01:08:23] they reap the benefits from all the
[01:08:25] early investments done by the government
[01:08:27] if you look across the globe if you look
[01:08:29] you know the europe india
[01:08:30] china do you see totally different ways
[01:08:33] to do it or do you think the u.s is so
[01:08:34] far ahead so it's really you know
[01:08:36] there's not so much
[01:08:38] alternative models we can learn from
[01:08:40] from other components
[01:08:42] well i think there's certainly things
[01:08:43] that we can learn from from every every
[01:08:45] country
[01:08:46] um because there are all sorts of
[01:08:48] different experiments that are
[01:08:49] effectively being conducted
[01:08:51] but that's a really important question
[01:08:52] because there are some key differences
[01:08:54] so
[01:08:54] the first observation that i would make
[01:08:56] is that the united states has
[01:08:57] traditionally
[01:08:58] been the largest funder of fundamental
[01:09:02] scientific research
[01:09:03] so nih has a budget of about 30 billion
[01:09:06] dollars
[01:09:07] every year that's the annual budget and
[01:09:09] it's been going on for decades so
[01:09:12] i think the us is very much far ahead
[01:09:15] but the other countries are catching up
[01:09:17] and particularly in asia china
[01:09:20] singapore they have made very big
[01:09:22] investments
[01:09:23] in the life sciences over the course of
[01:09:25] the last 10 years
[01:09:27] so i think that the balance of talent
[01:09:29] and power
[01:09:30] is shifting but even more important is
[01:09:33] the healthcare systems
[01:09:35] because the healthcare systems what we
[01:09:36] are willing to pay
[01:09:38] in terms of drug prices in terms of
[01:09:40] reimbursement policies in terms of
[01:09:41] coverage
[01:09:42] that very much determines how much
[01:09:44] innovation goes on
[01:09:46] and up until now the united states has
[01:09:49] been the most
[01:09:50] valuable market for any drug company to
[01:09:53] participate in so
[01:09:55] even countries that are in you know far
[01:09:57] away from us
[01:09:58] even those countries when they launch a
[01:10:00] drug they will typically
[01:10:02] launch it first here in the u.s because
[01:10:05] we are paying the highest prices now you
[01:10:08] know
[01:10:08] there's been a lot of debate about that
[01:10:09] right many people they say that that's
[01:10:11] not a feature
[01:10:12] that's a bug we should not be paying
[01:10:14] higher prices than people in canada
[01:10:16] or in the uk or in asia and this goes to
[01:10:19] a much
[01:10:20] deeper question that i am definitely not
[01:10:22] qualified to discuss but many
[01:10:24] health care economists have written a
[01:10:26] lot of very interesting things about
[01:10:27] which is
[01:10:28] what is the best health care system is
[01:10:31] it single-payer
[01:10:32] is it multi-payer is it some kind of a
[01:10:35] hybrid
[01:10:36] and i think that's where we have a lot
[01:10:37] to learn obviously the uk
[01:10:39] is a single-payer system canada
[01:10:41] single-payer
[01:10:42] germany multi-payer system the united
[01:10:44] states it's
[01:10:45] a free-for-all so we have very very
[01:10:49] different kinds of incentives that we
[01:10:50] are providing healthcare companies
[01:10:53] and that's something that i think can
[01:10:54] really use a lot of innovation
[01:10:56] and more research so can i come back to
[01:11:00] the
[01:11:00] illegal risk of a little startup or not
[01:11:03] a little one but
[01:11:04] essentially a biomedical startup there's
[01:11:07] also a risk that
[01:11:08] two more risks essentially that you
[01:11:10] cannot sell your company
[01:11:12] early enough how would you elaborate on
[01:11:15] that risk
[01:11:16] and there's also the risk that you know
[01:11:19] suddenly a competitor pops up and steals
[01:11:21] the market from you
[01:11:23] so how do you take this into account so
[01:11:24] if if you find more projects there might
[01:11:26] be a competing
[01:11:28] uh product coming and like your soccer
[01:11:30] players
[01:11:31] the third year third grade kids running
[01:11:34] together that discover suddenly the same
[01:11:36] thing or similar things so
[01:11:38] yeah so that that's a that's a really
[01:11:39] important question and there are several
[01:11:41] different questions
[01:11:42] underlying it so let me take a couple of
[01:11:44] aspects
[01:11:45] the first has to do with what happens if
[01:11:48] you you get interrupted in the middle of
[01:11:51] a project and
[01:11:52] uh or you can't raise money that is
[01:11:54] absolutely
[01:11:55] a huge problem that you can eliminate as
[01:11:58] you pointed out by having a longer term
[01:12:00] financing
[01:12:01] so for example if you finance your
[01:12:03] portfolio not by equity
[01:12:04] but by debt by long-term debt
[01:12:08] by 10-year debt by 20-year debt then
[01:12:11] you actually will not run into that
[01:12:13] problem but who would who would lend you
[01:12:16] money
[01:12:16] if you're not generating any cash flows
[01:12:19] so the key about debt financing has to
[01:12:21] be how to deal with that
[01:12:23] so that suggests that maybe a hybrid
[01:12:25] portfolio where you have late stage
[01:12:26] assets that are generating cash flows
[01:12:29] and early stage assets that will not
[01:12:30] generate cash flows for a decade
[01:12:32] a hybrid portfolio where you diversify
[01:12:36] the various different stages that you're
[01:12:38] in
[01:12:39] could be financed using these biobonds
[01:12:41] that's a possibility we don't know yet
[01:12:43] because it hasn't been done
[01:12:44] but i think that's a real issue but the
[01:12:47] overarching
[01:12:48] objective is we have a beta for a
[01:12:51] biotech company
[01:12:52] of two if we had
[01:12:55] financing that could take that biotech
[01:12:58] company from beginning
[01:12:59] all the way to end by end i mean fda
[01:13:02] approval
[01:13:03] if we could have financing for that
[01:13:05] entire
[01:13:06] process of clinical experimentation
[01:13:09] then the beta goes from two down to zero
[01:13:13] because then all that's left is just
[01:13:16] scientific risk right
[01:13:18] so this suggests that having a large
[01:13:20] enough portfolio
[01:13:22] having a longer term financing may be
[01:13:24] getting sovereign wealth funds
[01:13:26] that have much longer horizons than
[01:13:28] hedge funds
[01:13:29] or short-term financing maybe that
[01:13:32] will allow us to reach the full
[01:13:34] potential
[01:13:35] now the other thing that you mentioned
[01:13:37] was about biotech companies
[01:13:39] basically you know getting made obsolete
[01:13:43] by other technologies that's absolutely
[01:13:45] a risk that's part of innovation
[01:13:48] and rather than worry about that risk
[01:13:50] what we ought to be doing
[01:13:51] is to encourage more innovation by
[01:13:54] pooling the risk
[01:13:54] over a much larger portfolio so again
[01:13:57] portfolio theory and diversification can
[01:14:00] deal with that kind of risk
[01:14:01] and and by the way that risk is real to
[01:14:03] give you an illustration
[01:14:05] in the year 2000 if you took the list of
[01:14:07] the top 30 best-selling drugs
[01:14:09] the ones that generated the most revenue
[01:14:11] and you asked the question
[01:14:12] where did those 30 drugs come from
[01:14:14] pharma
[01:14:15] or biotech or academic medicine the
[01:14:18] answer
[01:14:19] is that 26 out of the top 30 drugs
[01:14:23] were developed by big pharma only four
[01:14:26] came out of biotech or academic medicine
[01:14:29] that was in the year 2000
[01:14:31] fast forward to the year 2018 which is
[01:14:34] the most recent year that we have data
[01:14:36] for this
[01:14:37] in the year 2018 the top 30 best-selling
[01:14:40] drugs in that year
[01:14:42] 26 of them came out of biotech
[01:14:46] or academic medicine 26. out of the top
[01:14:49] 10
[01:14:50] nine of them out of 10 came out of
[01:14:53] biotech
[01:14:54] and academic medicine so the balance of
[01:14:57] power has shifted
[01:14:58] from big pharma to uh biotech
[01:15:02] but it's not that big pharma is is lazy
[01:15:06] or or you know failing it is that big
[01:15:09] pharma's
[01:15:09] incentives their skills their objectives
[01:15:12] are focused on late stage
[01:15:15] drug development regulatory approval
[01:15:17] marketing distributing
[01:15:19] and licensing very good so thanks a lot
[01:15:23] uh andy it was a very insightful
[01:15:25] i learned a lot and i think the audience
[01:15:26] i also appreciate it but we went way
[01:15:28] over time
[01:15:29] i just want to conclude no it's uh
[01:15:32] that's good it's uh
[01:15:33] [Music]
[01:15:35] um we always conclude with a positive
[01:15:37] note
[01:15:38] so if you project into the future how do
[01:15:40] you see
[01:15:42] the future evolving let's say five or
[01:15:43] ten years how will the pharma industry
[01:15:45] how will biotech industry
[01:15:47] and how would the financing how do you
[01:15:49] envision it i mean you gave a clear
[01:15:51] vision but do you think it will happen
[01:15:52] and will we bring the ten years down to
[01:15:54] five years and this
[01:15:56] you know opens up another possibilities
[01:15:58] perhaps you can just end with one
[01:15:59] positive note we always end with a
[01:16:01] positive note
[01:16:02] uh at the end i absolutely think that
[01:16:05] we're going to see tremendous progress
[01:16:06] both on the scientific front and the
[01:16:08] financing front and the more financing
[01:16:10] that comes into the industry
[01:16:11] the faster is the scientific progress
[01:16:13] that we're going to make
[01:16:14] and believe it or not thanks to covet 19
[01:16:18] thanks to the pandemic i believe that
[01:16:19] government policy
[01:16:21] will now be much more focused on
[01:16:23] providing innovation
[01:16:24] and support for infectious diseases
[01:16:26] which was an area that was not getting
[01:16:28] enough attention prior
[01:16:29] to the pandemic so now i think we have a
[01:16:31] chance of being able to deal much more
[01:16:33] successfully
[01:16:34] with the next pandemic and with all
[01:16:35] sorts of health care issues
[01:16:38] thanks a lot andy i appreciate it hope
[01:16:41] to see you soon in the real world and
[01:16:43] let's stay in touch thank you so much
[01:16:44] it's been a pleasure and an honor
[01:16:46] and i wish you all good luck and please
[01:16:47] stay safe and healthy
[01:16:53] bye-bye