Scientist Stories: Andrew Lo, Can Financial Engineering Cure Cancer?

I'm proud to welcome you to this 4th annual Green Family Lecture Series. I'm Russ Cafflish. I'm the director of IPAM, the Institute for Pure and Applied Mathematics. I'm going to say a few words before turning it over to tonight's speaker. IPAM is a national math institute. Sponsored by the NSF, the National Science Foundation, and our mission is to promote the interaction of mathematics with other disciplines, everything from, from life sciences and medicine to engineering and physical sciences and even social sciences and humanities. We do this by bringing in visitors to workshops and long programs from not only from mathematics, but from a variety of other fields in our building that's over toward the center of campus. We have a current program now on new trends in financial mathematics, and there's a workshop right now on forensic data analysis as part of that long program. And this lecture also fits into that same theme. We really are, as part of our mission, We have public outreach, including public lectures like this, as well as other events. And we also invite you, if you'd like to get more involved, we have a mailing list that you can find out more things about that we're putting on, as well as a supporter society called the Frontier Society. The IPAM got started in 2000. It was created by Mark Greene and collaborators. Through a proposal to the National Science Foundation that was successful and has had great effects from there on. Mark was director of IPAM until 2008, and as part of this— as part of moving on, his family provided an endowment to support this lecture series, the Green Family Lectures. And this is, as I said before, this is the fourth in that series. We will, after today's talk, we will have a reception over in the Fowler Museum that's in the adjacent building. The theme of tonight's talk is very appropriate for IPAM. It's at the nexus of mathematics, finance, and biomedical research. I'm going to turn this over for the for the introduction to one of the organizers of this program, Renee Carmona. Let me first just say a couple of words about Renee. Renee is a professor at Princeton University, where he's the Paul Weitz Professor of Engineering and Finance in the Department of Operations Research and Financial Engineering. Renee has been a very influential figure in financial mathematics, where he started a journal part of SIAM and done other, many other activities, a SIAM activity group as well. He's also been recognized as a fellow of the Institute of Mathematical Statistics and of SIAM, the Society for Industrial and Applied Mathematics. So with that, I'm going to turn it over to René, who will introduce the speaker.

Thank you, but obviously it's not about me, it's about the speaker tonight. IPAM, our tax dollar at work, and working very well, actually. So the speaker tonight, Andrew Lo, is the Charles E. and Susan T. Harris Professor of Finance at the MIT Sloan School of Management. He's also the director of the MIT Lab of Financial Engineering. He has having a lot of influence through his textbook, for example. One of his most famous textbook written jointly with John Campbell and Craig McNeely, a textbook on the econometrics of the financial market has been used all over the country. I personally use it when I taught financial econometrics. But he had also a textbook with more catchy titles like Non-Random— sorry— Walk Down Wall Street, still with McKinley. And beyond all the books that he wrote, he also had a very strong influence on the academic community through student. So I was looking around, but I am sure that we have one of his PhD students, Mila, here today. And I am living and suffering every day at Princeton with one of his students, Yacine Haïssaliya, who is a colleague of mine and also a friend. So its influence is reaching far beyond his textbook. Andrew got his PhD in 1984 in economics from Harvard. Then after a short stint at Wharton, 3 years as an assistant professor and 1 year as a full professor, he joined the finance faculty at MIT in 1988. And I believe that he has been there since. Sorry. That's a trip to Chicago. He is extremely active in the community. He is currently the co-editor of the Annual Review of Financial Economics. He is the associate editor of many journals, including the Journal of Portfolio Management and the Journal of computational finance. Beyond that, he collected all the awards that we can think of. He's a Sloan Fellow, Guggenheim Fellow, and got the Paul Samuelson Prize. And he was the Financial Engineer of the Year. I forgot in which year. We're not going to talk about which year that was. But— and the institution, the financial institution, the academic institutions are all looking for His advice, they're all seeking Andy for lectures, advice, and of course the academic institution, I won't list them, but he's on many advisory board committees, SEC, CME, FINRA, and the New York Fed. His research has covered a broad range of, broad spectrum of applications, whether it's of economics or mathematics. More recently, he got interested in hedge funds and the shadow financial system, systemic risk, a reason why his advice is requested very often, and to the point of the talk, how clever a transition, to healthcare finance. So without any further ado, let's welcome Andrew Lo.

Well, I'd like to start by thanking Renée for that very generous introduction and want to acknowledge the Institute for Pure and Applied Mathematics at UCLA and Director Kaflish for sponsoring this event and inviting me to speak to all of you tonight. I also want to thank the Green family, Mark Green in particular, for this privilege. You know, we all appreciate donors that provide lecture series like this, but it's very unusual to have a donor who's also a colleague for whom you'd like to, you know, have their respect and, you know, comments on the work. So this is a very special treat for me. It's a bit of an unusual topic though, I have to say. As Renee mentioned, most of my research up until now has been focused on finance and investments. And so I have to start by a disclaimer, which is that I have no particular expertise in biomedicine. I'm not a healthcare economist, and up until a few years ago, I had absolutely no interest in any of these issues that I'm going to talk to you about today. The way I got interested was really for personal reasons. A number of friends and a family member were afflicted with cancer a few years ago. And in talking to them about their situation, I started learning a little bit more about not only their diseases but also various different therapies for dealing with cancer. And as a financial economist, I couldn't help but start looking into the financial aspects of the biomedical industry. And a very strange conundrum emerged as part of that investigation, and I'm not sure I fully understand it even today after having spent a fair bit of time thinking about it. So I'm going to ask for your help in thinking through this with me. So one of the things that I started with was to try to understand the biopharma industry. And I use that term very advisedly because the industry is not one industry. It's actually several different industries. But roughly speaking, you could think of it as biotech and pharma, biotech being the parts of the industry that focus on really the earlier scientific discoveries that come right out of the laboratory and into some kind of experimental and clinical development program. And then the pharma industry, which I suspect all of you know well, which is the big drug companies that basically take these ideas into clinical testing and ultimately FDA approval and marketing and distribution. So let me first start with a little bit of finance about these two industries and show you what the performance, the investment performance of the biotech and pharma industries has been for the last 20 years or so. So it turns out that from 1994 to 2015, the two industries have behaved very differently. What I'm going to show you in this plot is the cumulative return of the biotech and pharma indexes computed by the New York Stock Exchange, the NYSE ARCA indexes. So here's what would have happened if you had invested a dollar in the pharma industry starting from 1994 up until the present. That green line shows that cumulative return. During that same period, if you had invested a dollar and put it into the biotech industry as measured by the NYSE ARCA Biotech Index, this is what you would have gotten. Now, just looking at it, it seems like, gee, biotech is great, pharma not so great. But it actually hides a more interesting narrative that you have to see only through looking at the log scale returns. So when you look at this on a logarithmic scale, you notice a very interesting narrative that begins to emerge. And the narrative is that in the 1990s, the pharma industry was actually doing pretty well. So if you take a look at the slope of that line, that gives you a sense of the rate of return of that industry. So from the 1990s all the way to the early 2000s, pharma was doing quite well, and then somehow something changed, and it hit an inflection point where for a period of about 10 years, known as the lost decade in the pharma industry, pharma basically flatlined, and actually if you look at it very carefully, they ended up destroying shareholder value over that decade. Actually lost money, the pharma industry. This is an industry that most people would think have been making gobs of money, they actually ended up losing money for shareholders. They've done better since, so within the last 5 to 10 years they've made improvements and changed what they've been doing and so now they're back on a positive slope but not nearly as steep as the slope they were on in the 1990s. The biotech industry, on the other hand, has had a very different experience. Early on, biotech was pretty volatile, going up and down and up and then down, but since about the middle of 2003, biotech has been on a tear, very positive slope, and if you look at the most recent period, an even more positive slope. Now, there are many reasons for this. I don't begin to understand all of them, but I have to say it is curious that the inflection point in that time series for the biotech industry occurred right around April of 2003. In fact, I drew the line on April 14th of 2003 where it seems like right from that point there was a discrete shift and from that point forward the biotech industry has taken off even throughout the financial crisis which was 2007 to 2009. There was a little bit of a dip but it just kept on going. Does anybody know what happened on April 14th, 2003? No? But that's a good one. Yes? What about genomes? On that day, the US government announced that we had finally completed sequencing the human genome. And almost as if on cue, biotech industry jumps up and takes off. And as those of you who are part of that industry know, the sequencing of the human genome, bioinformatics, and big data have played an enormously important role in this industry. The biotech industry is at a new inflection point right now where discoveries are being made literally every week. Something remarkable is happening. I'll give you an example from my own home institution, MIT. A few weeks ago, a paper was published in Nature by a number of people. This is their typical author list of a paper in biology. And with the usual irony that the last author is actually the most important. The last author happens to be Manolis Kellis. He's a colleague of mine in the Computer Science and Artificial Intelligence Lab at MIT. Also a specialist in bioinformatics, biology, as well as computing. And in this article titled Integrative Analysis of 111 Reference Human Epigenomes, the paper describes various different on and off switches that turn genes on and off, but where they associate those genes with certain human conditions. Now, I have no idea what this paper says. I'm not qualified to read it. But I can look at the pictures. And if you look at the pictures— that's Manolis Kellis— if you look at the picture, here's a table that comes out of the paper. And I don't know what this means, but I can read the labels. And if you look at the labels on this table, you notice something kind of amazing. The labels are of human diseases. Crohn's disease, chronic lymphocytic leukemia, Graves' disease, celiac disease, rheumatoid arthritis. This paper is the beginning of our understanding of how our genes switch on and off and what that does with respect to human disease. Now, it's not nearly ready yet for prime time. We don't have drugs yet that deal with this. But if you talk to any biologist who's aware of these developments, they will tell you that this is an unprecedented time of discovery and development of really dramatically new therapeutics. So what's the conundrum that I referred to at the beginning? Well, the conundrum is this. We are at this inflection point where we're in the midst of these incredible breakthroughs. And yet, in the midst of these breakthroughs, we are systematically pulling money away from these discoveries and putting them into other activities, both on the public side and on the private side. So you all know that NIH funding is going down. Even before the sequester and even before you take into account inflation, which has been higher in biotechnology than in other areas, we've been decreasing our funding for these activities. But with those other phenomenon, sequester kicking in and other problems, the funding is way lower than what it ought to be if you take a look at the trend. From the previous decades. But the private sector is what I'm most focused on. I don't understand what drives the public sector, but the private sector, we've been pulling money away from the biopharma industry as well. Now, you might be surprised to hear that because over the last 18 months biotech seems really hot, right? Lots of new IPOs, lots of money seemingly going into biotech, but those are going into companies that were started 8 to 10 years ago and are finally now ready to be debuted in the public markets. Many people think that there's a biotech bubble that's about to burst, and most importantly, if you look at new companies being started by venture capitalists, the number of biotech VCs that are around has actually declined dramatically from 201 in 2008 down to 137 as of 2013. The number of Series A financings for biotech startups last year was among the lowest that it's been over the last decade. Why? I don't understand this as an economist. I see all these great breakthroughs, and yet at this very moment when we're on the cusp of dealing with some really amazing therapies, we're pulling money out and investing them in other things. And so I don't know the answer to this, but as an economist, I have to admit that to somebody with a hammer, everything looks like a nail. So this seems to me to be an economic problem. And the problem that I'm conjecturing is increasing risk and uncertainty. Now this is a very strange narrative. What I'm arguing is that over the course of the last decade, because of all these scientific breakthroughs, because we've gotten smarter, that's actually made the risk go up, not down. And it's very counterintuitive because I have to tell you that as an economist focusing on financial investments, it's exactly the opposite. The more you know about a financial investment, the smarter you are about a particular type of an investment, typically that means the lower is the risk. But not in biopharma. The more you know, the more we understand about the mechanisms of disease, the many more possibilities there are to investigate, each one of which costs money and could lead nowhere. And so the complexity is growing geometrically and therefore the costs are going up geometrically and it becomes much, much less attractive to be investing. So let me give you an example of this and to do that let's forget about biomedicine. I'm gonna just ask you to put your investment hats on and tell me whether or not you would be interested in the following investment I'm gonna produce. And actually to make the point, Do we have any MFE students in the audience? Show of hands? Okay, so we have a few investment students. I'm going to ask you in particular to tell me what you think about the following investment. So the investment opportunity requires $200 million. I need $200 million in cash in order to engage in the investment. You don't have to give me all $200, but just ask yourself whether you'd be willing to invest a small amount of your retirement funds, your children's college education fund, your parents' 401(k). Would you be willing to invest $3,000 in this fund? I need $200 million, and it's going to take me 10 years for this investment to bear fruit. And the probability of success is 5%, which means with 5% probability, I'll show a positive return. With 95% probability, I will show no return whatsoever. That's the proposition. How many of you would be willing to invest your parents' money or your children's money in this? Show of hands. Nope. Thank you. Thank you. Usually when I ask this question, nobody even bothers to ask, what is the rate of return? Occasionally, I will get a student that will raise his or her hand and say, excuse me, you didn't tell us what the rate of return is. So let me tell you, these numbers are the back-of-the-envelope numbers corresponding to what it takes to investigate a single anti-cancer compound from the preclinical stages all the way to FDA approval. It's not the cost of developing a successful cancer drug. That's, as of today, on the order of $2.6 billion. $200 million is the out-of-pocket cost of developing one compound from beginning to end. But of course, in order to find a successful drug, you've got to try many different compounds because we know that the probability of success is only 5%. But in the unlikely event that you're successful, it turns out that by today's standards, the earnings of an anti-cancer drug is about $2 billion a year. And by the time you get it approved, you will have about 10 years of patent life left. So what we're talking about is a cash flow that is $2 billion from years 11 through 20 in the unlikely event that you're successful, which at a 10% cost of capital, which is about what it is for this industry, amortizes to a payout of $12.3 billion in NPV if you're successful and nothing if you're not. Okay? So that means that you either have a 51% compound return on your $200 million— $200 million becomes $12.3 billion with 5% probability. That's a 51% compound return. Or minus 100% with 95% probability. For those MFE students, let me translate this into numbers that you can more easily appreciate. That corresponds to an investment that has an expected return of about 12% with an annualized standard deviation of 423.5%. Okay, now how many people would be willing to invest their parents' or their children's money in this? Show of hands. Okay, I got 2 takers. This is way too risky for any one of us to invest in. Nobody wants to do this. And the problem is that that's what pharma companies and biotech companies are faced with. They need these funds in order to develop these therapies, but they're extraordinarily risky. One of the things that I did in getting myself up to speed was to talk to a biotech company in Boston— in Cambridge, actually— about therapies they were developing. At the time, my mother was diagnosed with non-small cell lung cancer. And so a friend of mine who is well familiar with the biotech community was very kind to make that introduction. So I had the privilege of meeting with the Chief Scientific Officer of this very successful biotech company. They already had lots and lots of funding. They weren't looking for any more. They were closed and they were quite successful. They were getting ready to do an IPO. And the Chief Scientific Officer was very gracious to meet with me. He had no idea who I was, but he asked along the Chief Financial Officer who actually did know who I was. He actually used one of my books when he was getting his MBA. And so the two of them had lunch with me, and I asked them over lunch what I thought was a relatively innocent question. They were developing a number of different therapies, some of which were experimental therapies for dealing with non-small cell lung cancer. So I asked them, when you make your scientific agenda up and decide which priorities to focus on, does your source of funding have any impact at all, any influence on that agenda? And they looked at each other, laughed ironically, and turned to me and they said, influence? Our funding drives our scientific agenda. It drives it. And I got to tell you that from the perspective of the son of a patient, I was absolutely outraged by that answer. I was really offended. Because what does stock market volatility, interest rates, and Fed policy have to do with whether you can cure cancer via angiogenesis inhibition or immunotherapy? Nothing. But it drives, it drives their agenda. So at that point, I started thinking, well, wait a minute. If that's the case, then maybe we ought to start thinking a little bit more carefully about the financing. What if we could actually reduce this risk and make it more palatable for investors? Instead of doing one at a time, What if we did 150 at a time? Now, of course, in order to do 150, you do need, well, 150 times $200 million, you need $30 billion. Where are you going to get $30 billion? Well, as an economist, I have a very simple answer. Let's assume we have $30 billion. Now, bear with me for a minute. Suppose we do have $30 billion. And we had 150 independently and identically distributed cancer projects. Now, I'm going to come back and talk about that assumption. It is an important assumption. If I had 150 of these shots on goal, to use a hockey term, if I had 150 IID shots on goal, each with a 5% probability of success, what's the probability that I'll have 1 or 2 successes? We can calculate that as a simple binomial probability. And the answer is that the probability that I have at least 2 hits out of 150 IID trials with a 5% probability happens to be 99.59%. That's the probability that I have at least 2 successes out of 150 trials. Now, if I have 2 successes out of 150 trials, how much money have I made for my investors? Well, I've made 2 times $12.3 billion, or $24.578 billion. So if I am able to generate $24.5 billion in 10 years with 150 projects, that means that I can issue debt. How much debt can I issue without defaulting? I can issue up to $24.5 billion of debt that expires in 10 years without defaulting, which means that at a AA-rated yield of 2.45%, I get $19.3 billion of cash today. Almost two-thirds of my $30 billion I can fund by issuing AA-rated debt. Now how do I know it's AA-rated? Well, the rating has to do with the default probability. AA-rated debt typically defaults around 30, 40, 50 basis points. That's the probability of AA-rated defaults. What's the probability of default here? It's 1 minus 99.59%. That's how I got AA rating. If you now allow me to use all of the tricks of the financial trade, tranching, collateralized debt obligations, credit default swaps, all the things that gave us the financial crisis, I can do even better than this. Now you're asking yourself, is that really a good idea? I mean, after all, we did have the financial crisis. And I have to tell you that that's actually what gave me the idea to do this. It's not that these techniques didn't work, it's that these techniques, they worked way too well. That's why we got a crisis. But imagine if you took these very same techniques and applied them to things that you cared about, like curing cancer. Could you make an impact then? And the answer is maybe. I think you can, but there's a lot more that needs to be said. And I'll say a little bit about this. The first thing I want to say is I want to go back to that IID assumption, because clearly we know that things aren't IID. So just to be clear, the probability of having k or more successes for a binomial distribution with 150 trials at a 5% success rate When they're IID, this curve looks like that. And when k 2, the probability of at least 2 successes is going to be 99.59%. That's what I got here. If we now assume that these binomial— that Bernoulli trials are equally correlated, we then get a 10% correlation bringing that down to about 96% probability of at least 2 successes. If the correlation goes up, to 40%, now the probability of at least 2 successes is down to 70%. And if the correlation is as high as 80%, well, then you're looking at only a 30% probability of at least 2 successes. And at that point, forget about issuing debt. Nobody's going to want to buy your debt because the probability of default is way too high. So correlation matters. When I showed this to one of my pharma colleagues, He was amazed by this very simple analysis and said, you know, this is exactly what I've been telling my colleagues about what we need to change in pharma, but I wasn't doing this kind of analysis. I was using my daughter's third-grade soccer team as the point of argument. And I said, well, what do you mean by that? And he said, well, I'm the coach for that third-grade soccer team, and if you ever coached a kid's soccer, you notice that one of the challenges is that all the kids, they crowd around the ball. And you got to tell them, spread out. Don't go to where the ball is. Go to where the ball will be. And when you look at World Cup soccer, you see the difference, right? That's the lesson from basic portfolio theory applied to drug discovery. It's that if you all crowd around the same target, the risks go up. In which case you're not going to get the investment dollars because they're going to look at the portfolio and say, "Nah, I don't think so, too risky." You've got to diversify. Now, there are reasons why pharma companies do aggregate around similar therapeutics, but there may be ways of diversifying using financial means, and I'll talk about that in a minute. Now, let me ask the obvious question that all of you are thinking. And I'm going to start wrapping up with a couple of remarks about how we might actually implement this. The first question that all of you are thinking is, do we really need $30 billion? Because that's a problem if we do. In fact, I was kind of naive about this. I got the $30 billion literally out of the back-of-the-envelope calculation that I just did for you, nothing more sophisticated than that. But in more involved simulations and in papers that I've published in a number of journals, We've actually taken historical data from the drug— of the drug industry, drug discovery trials. We've actually calibrated the parameters of success and looked at the correlations given that historical data. And we've varied the parameters in a variety of different ways to try to get a sense of exactly how much money do you need. And the answer is, it depends. It depends on what your ultimate portfolio looks like. For cancer, it turns out that you do need on the order of tens of billions of dollars. Whether it's 30 or 10 or 15, it's not that critical. It depends on the typical type of cancer, how broad you want to get, what your targets are. But for cancer, billions of dollars is what you need. However, in a recent paper that we published just a couple months ago in Science Translational Medicine, we actually studied a real live portfolio of projects at the NIH. It's a portfolio of rare diseases. Just by way of background, According to the 1983 Orphan Drug Act, a rare disease is a disease that affects 200,000 patients or less. And so it turns out that there are many, many rare diseases that afflict various different individuals. And for those rare diseases, first of all, the most common of them are rare genetic disorders, single mutations that cause a missing enzyme. And in that case, by definition, they're going to be uncorrelated with each other. So the assumptions that we apply to doing that back-of-the-envelope calculation and where we don't make those same assumptions for the larger simulations, it turns out that for rare diseases, you only need a couple of hundred million to have an impact on these diseases. And the rates of return that we're estimating for a portfolio of a few hundred million is on the order of 22%. That's a pretty attractive rate of return that even hedge fund managers would be envious of in this day and age. So depending on your targets, $30 billion may be one kind of target, but a few hundred million may be other. And now let me give you the bad news. The bad news is there are certain diseases for which $30 billion is not nearly enough. And I'm talking in particular about Alzheimer's. Alzheimer's is challenging for a variety of different reasons. And the first that I got wind of this challenge was really in comparing what was going on in oncology with what's going on in Alzheimer's. So in 2014, does anybody know how many new cancer drugs were approved in that one year? Now, drug approvals don't happen all the time. It takes years and years and years, lots of hard work, and a number of drugs fail. Well before they get to the FDA approval process. So it's not a common thing for a drug to get approved. How many new cancer drugs did we introduce into the world in 2014? Anybody have a guess at that number? 0?

1?

5? 10? Very good guess. Actually, 10 is the number. 10 new cancer drugs in 2014, and between 2013 And last month, the number is 27 new cancer drugs in that 2.5-year period. How many new drugs for Alzheimer's? Zero. How about in 2013? How many new drugs in Alzheimer's? Zero. How about 2012? How many? Zero. In fact, if you keep going all the way back, you've got to go all the way back to 2003. It's been over a decade since we've had a single new drug for Alzheimer's. Over a decade. Now why? Is it because Alzheimer's is a rare disease? It doesn't affect that many patients? 5 million Americans suffer from Alzheimer's. Is it because Alzheimer's is not a particularly costly disease? Well, Medicare and Medicaid this year will spend $200 billion reimbursing patients for their medical expenses for Alzheimer's, not the care costs, but literally drugs and doctor's visits. $200 billion this year spent on Alzheimer's. It turns out that developing Alzheimer's drugs is really hard. In the paper that we published last year in Science Translational Medicine, we showed that the assumptions that I gave you for cancer, they don't apply to Alzheimer's. It doesn't take 10 years to develop an Alzheimer's drug, it takes 13. And it's not $200 million to develop an Alzheimer's drug, it's more like $500 to $600 million. And in terms of the probability of success, well, let's just say that it's less than 5% because we actually don't know what the number is since we haven't had a success in 10 years. And so when you factor all of that in, it turns out that by our estimates, in order to bring the risk down, you need to have $100 billion But the problem with $100 billion is that that requires something like 180 targets. We don't have 180 targets for Alzheimer's. We have right now 2: beta amyloid and tau. Those are the 2 hypotheses that people are prosecuting for developing Alzheimer's therapeutics. And the problem is that these are correlated, which means that you're not going to get much risk reduction, which is why why we don't see a lot of new drugs. It's because it's just not worth it in terms of risk-reward trade-off. It's a terrible thing to say given the suffering that Alzheimer's imposes on our parents and grandparents, but that's the way it is. The economics don't work. And part of that is because the basic science is not as well developed. We have lots of targets for cancer because the National Cancer Act in 1971, when we declared war on cancer, caused lots of new scientific innovations to develop the targets, including the Human Genome Project. We don't yet have that for Alzheimer's, and the private sector will not do this. So the question about how much money we need really depends upon what our objectives are. For some objectives, we don't need a lot of money. For other objectives, even an ocean of money is not going to help unless we take that money and invest it in the right ways. By the way, if you're interested in running any of these simulations, we've actually coded up all of our simulations in both MATLAB and R. We put it on our website with an open-source license, anybody to take it, steal it, copy it, do whatever you like, try it out and see whether or not you can make heads or tails of these parameters. We're not experts in the literature. We're using data that we've gotten from the industry, so we encourage others to experiment with it. The last couple of points I want to make, and I'm happy to open up to Q&A so that we can get the discussion going. The last two points is, one, isn't pharma already doing this? And two, how do we go about getting this going? Is it really realistic to try to raise funds to do this? So let me first deal with the issue, isn't pharma already doing this? The answer is, sort of. They used to be doing this a lot more, But over the course of the last decade or so, they're doing less and less of it. And part of the problem is because the risks are going up, and as publicly traded companies, pharma is being told not to do this. So let me give you an example. Pfizer is one of the largest pharma companies in the world, the largest US drug company. If you take a look at Pfizer's balance sheet, you notice something very interesting. Pfizer, as of 2014, had $36 billion of cash just lying around, sitting on its balance sheets. Now, $36 billion, that looks an awful lot like one of these funds that I was talking about, these mega funds for developing cancer therapeutics. They don't just focus on cancer, of course. So that $36 billion looks awfully tempting, and especially when you realize that of that $36 billion, $31.5 billion is actually from long-term debt. They've issued bonds to get that $36 billion. $31.5 is debt. So Pfizer is a mega fund, isn't it? Except that what they're doing with their cash is not investing in early-stage drug therapeutics. What they're doing is looking around for opportunities that they can invest in. In other words, you know that Pfizer is trying to buy AstraZeneca, that deal didn't work out, so they're on the shopping, on a shopping trip to look for other companies. By the time a drug company gets to the point where they're publicly traded, they actually have to respond to shareholder demands. And one of the things that shareholders are telling the pharma companies is cut risk, focus on what you're good at. And it turns out that the pharma business is really two businesses. One is developing new drugs and the other is getting drugs FDA approved and marketing and distributing them. It turns out that the pharma companies are very good at that second aspect of the business. But most people are not that good at the first part because it's hard, it's really hard to develop these drugs. It's not a criticism of pharma at all. And so, let me give you an example of how we've been telling these pharma companies. In 2010, Morgan Stanley issued a report titled Exit Research and Create Value. That was its advice to the pharma industry. And in that report, this analyst pointed out the solution: our economic value-added analysis supports replacing research with search. On current market economics, we estimate that a dollar invested in in-licensed compounds— that means when you license it from somebody else outside in— a dollar invested in in-licensed compounds will on average deliver 3 times as much value as a dollar invested in in-house research. In other words, get out of the business of early-stage drug discovery and focus instead on acquiring companies after they've de-risked their portfolio of compounds and focus instead on what you're good at, which is taking those compounds all the way through FDA approval and marketing and distributing the drugs. And you know what? Pharma has been listening. So this is just in 2012, the number of job cuts that have been occurring in the pharma industry. This has been going on now for about a decade. And remember I showed you that graph at the very beginning of the pharma companies and that lost decade and then getting better? The way that they're getting better and improving their share price is by listening to Morgan Stanley and cutting their R&D expenditures and focusing instead on late-stage drug development and approvals. So here's an example, a classic example. Just a few weeks ago in Boston, Sanofi announced that they were partnering with a biotech company called Voyager Therapeutics. And they're going to be spending upwards of $845 million by partnering with Voyager on a particular set of drugs. That same day that they made this announcement, they also announced that they were laying off 100 R&D people in Boston. And this is how shareholder value gets created. It's not a bad thing or a good thing. It's just a thing. This is what's been going on. And so the pharma industry has been doing less and less of it because it's harder and harder, and they don't know where the value is going to be. And so instead of trying to come up with those early-stage insights, let's focus instead on what we're good at. Now, if you're skeptical about this narrative, as I was, I started looking around at other industries just to get a sense of whether or not this has been playing out in those domains. And I've got to tell you that there is an exact parallel right here in Los Angeles. And the exact parallel is with the film industry. It turns out that the film industry is not actually one business. It's two businesses. A typical film studio is involved in two different activities. One, One is in making movies. The other is in marketing, licensing, distributing movies. That sound familiar? It turns out that making movies is actually quite expensive and producing blockbuster films is quite hard. But licensing, marketing, and distributing movies, that's a very good business. In the parlance of finance, it's a very high Sharpe ratio business. High return, low risk. And so it turns out that these two different businesses actually have very, very different risks. So there was a paper that was written by a couple of authors, Arthur Devaney and David Walls, about uncertainty in the movie industry. Can you quantify the risks of producing a blockbuster film? And the answer is an answer that was given years ago by the famous screenwriter William Goldman, who in a book on the film industry wrote a very simple phrase that summarized it, which is, "Nobody knows anything." Nobody knows how to produce blockbusters. And so these authors, they looked at the probability distribution of the rates of return of films. And what they find is a big spike up here. There are a few films that earn more than 100% in terms of their IRR. And then there's a tail over here And then sort of mess in the middle. Nobody knows anything. In fact, the probability of a blockbuster is about the same as the probability of producing an anti-cancer drug. It's about 5%. And these two authors, they actually looked at all sorts of factors, like whether or not you have a big-name movie star, whether you've got a great scriptwriter, whether you've got a great director. Doesn't matter. Nobody knows anything. There's no way to tell. Whether or not you're gonna get a blockbuster. And so what's happened over the course of the last 10 or 15 years is that the movie studios have gone to a very different model, the first being DreamWorks SKG, founded by Steven Spielberg, Jeffrey Katzenberg, and David Geffen. What they did was they did a deal. They did a deal that said, you know what, we're gonna borrow money from bondholders, and what we're gonna do is to use the money to finance our movie production. So instead of investing in one movie or another, you're going to invest in a portfolio of movies that we're going to produce. And you give us the money, we'll give you an interest rate, and we're going to de-risk that portfolio by being able to do many, many of these, and we do financing. That's how they got started. 2005, Gun Hill Road, a deal, a hedge fund that was set up to finance movies from Sony and Universal. $600 million for a portfolio of 17 pictures. 2005, Legendary Pictures did a deal with Warner Brothers where they put up $500 million for 25 pictures, which included Batman Begins and The Hobbit and so on. From 2005 to 2008, $12 billion was financed using these kinds of securitization techniques, $4 billion of that from hedge funds who had previously never invested in movies. They're not in the business of making movies. They're in the business of making reasonable rates of return. They can do that with the proper kind of financing. So I want to argue that new business models are emerging in the healthcare industry using these kinds of creative financing tools. And part of the reason that these new business models are emerging is because the traditional models are broken. They are not able to deal with the kinds of risks that are involved. We need a new business that's not a pharma company, not a biotech company, not a VC, not a mutual fund, but some amalgam of all of these things because if you look at the drug development process, it turns out that the risks and rewards are very different across each of these stages and that means that the natural investors are different, which means that the financing methods have to be different. You need to think more carefully about the financing. You need to let the financing science drive the financing rather than having the financing drive the science as it did in that biotech company that I described. The mega fund structure is capable of doing that. And we're actually seeing some beginnings of this. So just about 4 weeks ago, 5 weeks ago, we actually had an interesting event occur. A company in Ireland, a group of ex-pharma company executives, they decided to launch a fund to invest in early-stage life sciences companies. But instead of going the usual route of creating a VC fund, they did an IPO. They did an IPO to raise money to invest in funds and companies that they had not pre-specified. In fact, their business model was, the plan is to acquire majority or significant minority equity positions in private PIPO pre-trade sale operating businesses businesses in the life sciences industry. That's it. Trust me, I'm gonna invest, just give me money. They raised €330 million. Now, the government did help, the Irish government invested €50 million of those €330 million, but nonetheless they actually found incredible appetite among investors because they were able to argue that by constructing a portfolio we can reduce the risk. So let me start wrapping up by asking you a very simple question here and proposing a potential answer. Can California cure cancer? I'm going to argue that with enough imagination, the answer is yes, absolutely California can cure cancer. And let me give you a little bit of a sense of what that imagination might look like. So all of you know better than I do the wonderful academic medical centers, research facilities, hospitals, and researchers in the state of California. Some of the most exciting things in biotech are happening right here in this state. If we created a portfolio of all of these projects— now, I'm not arguing that you should keep it to within California, but let's suppose you did that. If you did UCLA, USC, Caltech, City of Hope, UCSF, you brought all together these various different projects that are going on and you put them into a single portfolio. Would you invest in that? Well, let's suppose that the portfolio was going to be managed by an incredible advisory board of experts. For example, on the scientific front, you had David Agus, David Baltimore, Mark Davis, Jennifer Doudna, Charles Homsey, and Frank McCormick. These are all California residents. On the business side, you had Eli Broad, Ron Lin, Elon Musk, Gordon Moore, Dick Roll, Eric Schmidt, and Bill Sharp. Suppose you had these people involved in managing this portfolio. Now, you might say, well, wait a minute, how are you going to get these folks? These are very busy, high-profile individuals. The folks in blue, they have no idea who I am. If I called them, I doubt they would return my phone calls. The folks in green, most of them, they do know who I am, which is why I know they won't return my phone calls. But if you raised $30 billion, Every single one of these individuals will return your phone calls, not because of the money. The folks in blue, they're not motivated by money, they couldn't care less about the money. And the folks in green, they have all the money. But if you raise $30 billion, you can actually make a difference. And my guess is that every one of them will wanna be part of that. So if that were the structure and we had a large, reputable, California-based mutual fund complex offering such a fund with these people with the target of curing cancer of all types. How many of you would invest $3,000 of your 401(k) as a one-time investment in this portfolio? Now before you answer, I'm gonna ask all of you to answer. Most, most of you would not invest in my $200 million project, but I'm gonna ask how many of you would invest in this. Before you answer that question, I'm gonna leave you with one final story about why it is that you might consider investing. And the story has to do with a colleague of mine at MIT by the name of Harvey Lodish. Harvey Lodish is a cell biologist, a faculty member at MIT and at the Whitehead Institute. And I got to tell you that I want to be Harvey Lodish. And let me explain why I want to be Harvey Lodish. So Harvey came to MIT years ago and in 1986, he was approached by a venture capitalist who who is seeking some help in developing a drug for a very rare disease called Gaucher disease. It's a rare disease that affects maybe 10,000 patients or maybe at this point 50,000 patients in the U.S. It's a genetic disorder that causes a missing enzyme that needs to be replaced. So he was hoping to develop an enzyme replacement therapy and Lodish, a cell biologist who was a specialist in the behavior of cell membranes, had a critical piece of scientific information that he could contribute. So he agreed. And from 1986 to 1992, he worked with this fellow to create a therapy, an enzyme replacement therapy that was approved in 1992. Great story. And the drug was approved. They started a company. The company you may have heard of. It's called Genzyme. Quite successful. Harvey did well. The founders did well. The shareholders did well. The patients did well. And everybody was happy with the outcome. But that's not the end of the story. In 2002, Harvey's daughter became pregnant with her first child, Harvey's first grandchild, and this child was diagnosed in utero with Gaucher's disease. Now, this disease doesn't manifest itself until puberty, and so there's a chance that you might have the gene for it, but it doesn't emerge. Well, in 2012, Harvey's grandson Andrew did develop Gaucher's disease, and he's doing just fine thanks to Grandpa's drug. When Harvey developed that drug, his daughter didn't even have children. He wasn't doing it because of his grandchildren. And this is why I want to be Harvey Lodish. I want to be able to do this for my grandchildren, but I can't. I'm not a molecular biologist or an MD-PhD. I don't know how to do that.. But I can if I invest in a fund that develops cures for my grandchildren. So now let me ask you, how many of you would be willing to invest $3,000? Well, I only needed 8% of you to raise your hand because 8% of 130 million households gives me $30 billion. The money is here. You saw it for yourselves. It's not that we don't have the money. We have the money. We don't have the vehicle to channel the money into the right resources. That's a finance problem. Now I know that finance is not the most popular subject even among our MFE students. Finance has gotten a bad rap, maybe deservedly so, but we want to be careful not to throw out the baby with the bathwater. It's a little bit, you know, blaming finance for the financial crisis is a little bit like blaming real numbers and arithmetic for accounting fraud. Yes, they're related and it's part of the system, but you've got to ask a little bit of a deeper question about what went wrong. Finance doesn't have to be a zero-sum game. It is possible to do well by doing good if we find the right structures. And so instead of eschewing finance, let's think about better ways of using finance to be able to deal with some of society's biggest challenges. And with your help, I think we can do it. There's a role for mathematicians, statisticians, big data, financial engineers, and biomedical experts. But the only way we're gonna do that is by collaborating, which is one of the reasons I'm so grateful for the opportunity to speak here at IPAM. It's because, for those of you who are still students, based on pure geog— biology, you are the future of the financial industry. So, by taking some of these ideas and thinking about how to create solutions to these problems, we're going to actually have a much better financial system for the future.

Thank you. We have time for questions.

Since we're recording this, we just ask that You mentioned that in our area of big pharma that people were being advised to do away with research and to reward their stockholders with lower-risk investments. Is that the case for every nation? So, for example, Is that happening in China? Is that happening in India? And my second question is, did you figure out what the mathematical chances are for the researcher to have been the same person to have had a grandson with this rare disease?

So the short answer to the second question is no, I haven't, but, you know, there are statisticians out there that have done similar calculations. Percy Diaconis, of course, did a calculation of, you know, winning the lottery twice in a year, and Mosteller and Diaconis have done a number of these coincidence type of calculations. It's high, and I would argue it may be even higher to think about investing in drugs and ultimately having your grandchildren benefit from those investments. I think that'd be a cool calculation to do. With regard to your first question, first of all, I want to make it clear that pharma is still doing a lot of research, so I'm talking about trends, not about actual current situation. Pfizer spends billions of dollars every year on research. It's just that the kind of research that they're doing and the trend is going down, not up. They're focusing on later stage, and you can understand why that's the case. Other countries, I'm not as familiar with India. I know that in China, they're spending a lot of money on genomics. The state government is spending a lot of money on genomics. But they still don't have the same private sector infrastructure that we have here in the United States. So they're in early days, but I heard from Eric Lander at a talk that he gave recently that China is actually generating a lot more genomic data than we are here in the U.S. They are probably the single— the number one producer of genomic information, and so I think that there's definitely a competitive angle that we ought to consider. We could easily be left behind in an industry that the United States basically pioneered. So I think that's a very important lesson to learn. I should also mention that, you know, pharma is not standing still. They understand these trends and they're engaging in a lot of creative ways of dealing with it. Some of those ways you might not like. For example, I don't know how many of you have heard of this pharma company called Valiant Pharmaceuticals. So Valiant is a company that's an incredible success story. It was a pharma company from quite a few years ago, but in 2009 an investment banker, a consultant who used to work at McKinsey, a fellow by the name of Mike Pearson, took it over. And over the course of the last, you know, 7 or 8 years that he's been heading up that company, he's produced remarkable, remarkable investment returns for his investors. And the way he's been doing it is by focusing on exactly this portfolio theory idea, by cutting R&D expenditures and instead Instead, taking the money and acquiring companies and putting together portfolios and then raising the prices of drugs to be able to take advantage of these kinds of market conditions. So just to give you a sense of that, the typical pharma company spends about 20% of its budget on R&D, 20% of its earnings or sales on R&D. Valiant spends 3% on R&D. Companies engage in lots of transactions. Well, last year, 2014, Valiant engaged in 40 M&A transactions, 4-0. So they're an example of what new pharma might look like, and it's disturbing to a number of people who feel that pharma should be about developing drugs. The problem is that the economic realities don't necessarily point in that direction, that the biotech industry may be the new engine for developing drugs that later get marketed and distributed by the later-stage pharma company efforts. Do you think there's a role for university endowments, pension funds, sovereign wealth funds to participate in these or build a mega fund? There's a huge role. I think that those institutional investors obviously played a big role in mortgages and mortgage-backed securities. They didn't traditionally invest in these kinds of asset classes. You didn't have pension funds lending money to people in Stockton, California to build homes, but they did that as part of this whole financial innovation, and you could argue that that wasn't a good thing because we went overboard. Nonetheless, there are millions of Americans today that are in homes that they have not defaulted on, that have not been foreclosed on, and they otherwise couldn't have afforded without these kinds of structures. So I think we have to be careful not to discard the lessons that we learned with the financial crisis, we need to make use of them in ways. So I think that pension funds, sovereign wealth funds, foundations, endowments, all can play a huge role, particularly because in many cases they have much longer horizons. So they're not looking for quick milestones 3 to 5 years out. They may be willing to invest for 20 or 30 years. In fact, MIT— this blows my mind— MIT, a few years ago, we issued 100-year bonds. It pays in 100 years. It pays interest along the way, but the principal you don't get back for 100 years. And that bond issue was oversubscribed, like 2 to 1. I think we ended up issuing $700 million of those bonds at an interest rate of 5.623%. Think about that kind of horizon and what that would do to the drug development process, how you can now allow the financing to actually match the horizon of the scientific endeavors.

That's huge.

Well, no more questions.

Let's thank Andrew again.

Thank you.