When people in the biopharmaceutical world speak of the valley of death, they’re not talking about a geographic hot spot where a lethal disease has infected a large number of people. Rather, the valley of death is where scientists in a research program have spent the last of their grant money, or a young biotech company has burned through its preliminary financing, and, though a program may still be promising, potential funders decide that they don’t want to risk the costs of taking it to the next level, and it dies. Currently, the valley of death, and similar financial constraints, are hampering vaccine development, which health officials are counting on to release us from the scourge of COVID-19 and, perhaps even more important, from future pandemics that the novel coronavirus portends.
Judging from the Times’s “Coronavirus Vaccine Tracker,” a lack of research funds wouldn’t seem to be a serious problem. More than a hundred and sixty-five vaccines are in development around the world. Seven, including three in China, are in advanced human trials. The two most talked-about in this country—Moderna’s, in Cambridge, Massachusetts, and AstraZeneca’s, at Oxford University—rely on new technologies, based on genetic engineering, being deployed with seemingly unprecedented speed. Similar programs run by Pfizer, Johnson & Johnson, and Novavax claim to be not far behind. All speak of a vaccine by 2021. The U.S. government is paying up to ten billion dollars in subsidies to five of these programs, for trials and manufacturing, under a plan called Operation Warp Speed. Just three weeks ago, Pfizer was awarded a contract worth nearly two billion dollars, to produce up to six hundred million doses. On Tuesday, it was announced that Moderna would supply a hundred million doses for a billion and a half dollars.
This would all suggest that governments and Big Pharma are doing all that can be done. And that may well have been the case if the vaccines currently under development were in response to a virus so novel that no one could have anticipated it; or if nothing like it will ever occur again, so that emergency funding on this scale will never be called for again, either. But neither is true. We have known about other fatal coronaviruses, such as SARS and MERS, for years—and we have also known that they would not be the last that we encountered. Moreover, vaccines are developed in distinct research platforms: some work with genes, others with proteins, or parts of the virus itself. Each requires particular methods of development and manufacturing, and those methods, not to mention the extensive testing regimes that are required—in this country, by the Food and Drug Administration—all make the typical financial pitfalls of drug research worse in the case of vaccines.
For a start, costs are higher. You aim to vaccinate millions of people who are still healthy, and whose responses to an inoculation may vary greatly by age, gender, body type, and other factors. You don’t want to harm anyone unnecessarily, so advanced trials may involve tens of thousands of people. Then there’s the issue of demand. A drug that could successfully treat a chronic disease such as cancer would be in demand around the world every day. Right now, so would a COVID-19 vaccine, but that’s not typical of vaccine markets; indeed, vaccine producers will likely face radically shrinking markets as soon as their products become widely available. Other vaccines, for diseases such as malaria, which target poor countries, are often purchased by philanthropic organizations for distribution, and do not promise satisfactory returns.
SARS and MERS both subsided before vaccines for them could attract investment. But, had investments been made, Ray Jordan, Moderna’s chief corporate-affairs officer, told me, by now we might have at least learned what levels of antibodies were sufficient to provide immunity against other coronaviruses. He added that this knowledge might have expedited advanced human trials for COVID-19, and, just as crucially, that we might now have had “a vaccine-manufacturing system and supply-chain up and running.”
Little wonder that, prior to COVID-19, only four of the Big Pharma firms—most prominently, perhaps, Britain’s GlaxoSmithKline—considered the market for vaccines attractive enough to commit to developing them in their own labs. Pfizer’s COVID-19 vaccine, for instance, is being developed by BioNTech, a twelve-year-old German biotech firm with which it entered into an agreement in August, 2018, to do research on cancer and infectious diseases, where the former promised a more profitable return. As things stand, even at the current accelerated pace, it will take a year, at least, from the time the coronavirus genome was sequenced, in January, for a COVID-19 vaccine to reach the first of us. And who knows what new mutations or viruses may present themselves as we wait?
Back in 2005, Anthony Fauci, who was already the director of the National Institute of Allergy and Infectious Diseases, presented the problem in a lecture to the Milbank Memorial Fund. “Infectious diseases accounted for about twenty-six per cent of the fifty-seven million deaths worldwide in 2002,” he said, deploring the persistence of HIV/AIDS, malaria, and tuberculosis. But, “faced with the choice of putting two hundred million dollars into a new area, will pharmaceutical companies make a product to combat an emerging microbe, for which there is an uncertain market, or will they develop a new Viagra or a better Lipitor?” No indictment was implied, at least not of individual companies, which have to answer to shareholders; the companies were behaving normally, which only makes the crisis more vexing. The challenge, Fauci suggested, calls for new approaches to producing vaccines not just for a moment of crisis but as part an ongoing, long-term process, and not just in the United States but internationally—in short, a radical breakthrough not only in biological science but also in financial engineering.
Andrew W. Lo and Roger M. Stein have been making the same argument for much of the past decade at the Massachusetts Institute of Technology’s Laboratory for Financial Engineering. Lo, the lab’s founding director, was born in Hong Kong in 1960, and was raised by his mother, in Queens. He did a Ph.D. in economics at Harvard, research in advanced financial models at Wharton, and, by 2010, he had become an innovator in managing large asset funds. He was one of the original quants, applying advanced mathematics to devise complex financial products. Then, Lo told me, “Over a four-year period, six people that were close to me all died of cancer, including my mother. I thought that, somehow, knowing me was carcinogenic.” He couldn’t reconcile himself to the slack pace of finding new drugs, and, he said, “The more I studied this, the more I realized that finance actually plays a huge role in drug development—in many cases, way too big a role.” It also became clear to him that it would be possible to secure better, more efficient financing “just by taking some of the tools that we use routinely in financial-portfolio management and applying it to drug-portfolio management.”
Drugs are a risky business and, for equity investors hoping to eventually share in the profits, each stage of development presents an escalated risk. Lo reasoned that substantially lowering the risks, even if it meant correspondingly lowering the rewards, could attract investment instead from ordinary bond markets—that is, from managers of pension funds, university endowments, and sovereign-wealth funds, who control a great deal of money and generally invest in low-risk, low-return assets. He took the idea to Stein, an expert in risk management and machine learning whom he knew from various professional encounters. Stein currently teaches at New York University’s Stern Department of Finance, but at the time he was the president of the research lab at Moody’s, the bond-rating agency. As it happens, he had also been thinking about applying risk management to medical science, because his father, too, had cancer. “I thought at the time it was going to be a quick project, because the math didn’t sound complicated to me,” Stein said. “Had I understood how many different pieces have to come together for drugs to get developed—all the things I had to learn—I wouldn’t have thought it was going to be so quick.” (Their model was first published in 2012, in Nature Biotechnology, in a paper titled “Commercializing biomedical research through securitization techniques,” which they co-authored with José-María Fernández, who is now at Altamar Credit, in Spain. It was refined in a 2019 paper, “Funding Long Shots,” co-authored with the derivatives expert John Hull, of the University of Toronto, and published in The Journal of Investment Management.)