Recent Research

Culture is a potent force in shaping individual and group behavior, yet it has received scant attention in the context of financial risk management and the recent financial crisis. I present a brief overview of the role of culture according to psychologists, sociologists, and economists, and then present a specific framework for analyzing culture in the context of financial practices and institutions in which three questions are answered: (1) What is culture?; (2) Does it matter?; and (3) Can it be changed? I illustrate the utility of this framework by applying it to five concrete situations—Long Term Capital Management; AIG Financial Products; Lehman Brothers and Repo 105; Société Générale’s rogue trader; and the SEC and the Madoff Ponzi scheme—and conclude with a proposal to change culture via “behavioral risk management.”

Risk aversion is one of the most basic assumptions of economic behavior, but few studies have addressed the question of where risk preferences come from and why they differ from one individual to the next. Here, we propose an evolutionary explanation for the origin of risk aversion. In the context of a simple binary-choice model, we show that risk aversion emerges by natural selection if reproductive

risk is systematic (i.e., correlated across individuals in a given generation). In contrast, risk neutrality emerges if reproductive risk is idiosyncratic (i.e., uncorrelated across each given generation). More generally, our framework implies that the degree of risk

aversion is determined by the stochastic nature of reproductive rates, and we show that different statistical properties lead to different utility functions. The simplicity and generality of ourmodel suggest that these implications are primitive and cut across species, physiology, and genetic origins.

Despite many compelling applications in economics, sociobiology, and evolutionary psychology, group selection is still one of the most hotly contested ideas in evolutionary biology. Here we propose a simple evolutionary model of behavior and show that what appears to be group selection may, in fact, simply be the consequence of natural selection occurring in stochastic environments with reproductive risks that are correlated across individuals. Those individuals with highly correlated risks will appear to form “groups”, even if their actions are, in fact, totally autonomous, mindless, and, prior to selection, uniformly randomly distributed in the population. This framework implies that a separate theory of group selection is not strictly necessary to explain observed phenomena such as altruism and cooperation. At the same time, it shows that the notion of group selection does captures a unique aspect of evolution—selection with correlated reproductive risk–that may be sufficiently widespread to warrant a separate term for the phenomenon.

Using account level credit-card data from six major commercial banks from January 2009 to December 2013, we apply machine-learning techniques to combined consumer-tradeline, credit-bureau, and macroeconomic variables to predict delinquency. In addition to providing accurate measures of loss probabilities and credit risk, our models can also be used to analyze and compare risk management practices and the drivers of delinquency across the banks. We find substantial heterogeneity in risk factors, sensitivities, and predictability of delinquency across banks, implying that no single model applies to all six institutions. We measure the efficacy of a bank’s risk-management process by the percentage of delinquent accounts that a bank manages effectively, and find that efficacy also varies widely across institutions. These results suggest the need for a more customized approached to the supervision and regulation of financial institutions, in which capital ratios, loss reserves, and other parameters are specified individually for each institution according to its credit-risk model exposures and forecasts.

Unlike other industries in which intellectual property is patentable, the financial industry relies on trade secrecy to protect its business processes and methods, which can obscure critical financial risk exposures from regulators and the public. We develop methods for sharing and aggregating such risk exposures that protect the privacy of all parties involved and without the need for a trusted third party. Our approach employs secure multi-party computation techniques from cryptography in which multiple parties are able to compute joint functions without revealing their individual inputs. In our framework, individual financial institutions evaluate a protocol on their proprietary data which cannot be inverted, leading to secure computations of real-valued statistics such as concentration indexes, pairwise correlations, and other single- and multi-point statistics. The proposed protocols are computationally tractable on realistic sample sizes. Potential financial applications include: the construction of privacy-preserving real-time indexes of bank capital and leverage ratios; the monitoring of delegated portfolio investments; financial audits, and the publication of new indexes of proprietary trading strategies.

Moderator Martin Leibowitz asked a panel of industry experts—Andrew W. Lo, Robert C. Merton, Stephen A. Ross, and Jeremy Siegel—what they saw as the most important issues in finance, especially as those issues relate to practitioners. Drawing on their vast knowledge, these panelists addressed topics such as regulation, technology, and financing society’s challenges; opacity and trust; the social value of finance; and future expected returns.

Economic shocks can have diverse effects on financial market dynamics at different time horizons, yet traditional portfolio management tools do not distinguish between short- and long-term components in alpha, beta, and covariance estimators. In this paper, we apply spectral analysis techniques to quantify stock-return dynamics across multiple time horizons.Using the Fourier transform, we decompose asset-return variances, correlations, alphas, and betas into distinct frequency components. These decompositions allow us to identify the relative importance of specific time horizons in determining each of these quantities, as well as to construct mean-variance-frequency optimal portfolios. Our approach can be applied to any portfolio, and is particularly useful for comparing the forecast power of multiple investment strategies. We provide several numerical and empirical examples to illustrate the practical relevance of these techniques.

Technological advances in telecommunications, securities exchanges, and algorithmic trading have facilitated a host of new investment products that resemble theme-based passive indexes but which depart from traditional market-cap-weighted portfolios. I propose broadening the definition of an index using a functional perspective—any portfolio strategy that satisfies three properties should be considered an index: (1) it is completely transparent; (2) it is investable; and (3) it is systematic, i.e., it is entirely rules-based and contains no judgment or unique investment skill. Portfolios satisfying these properties that are not market-cap-weighted are given a new name: “dynamic indexes.” This functional definition widens the universe of possibilities and, most importantly, decouples risk management from alpha generation. Passive strategies can and should be actively risk managed, and I provide a simple example of how this can be achieved. Dynamic indexes also create new challenges of which the most significant is backtest bias, and I conclude with a proposal for managing this risk.

Importance  Few patient populations are as helpless and in need of advocacy as children with cancer. Pharmaceutical companies have historically faced significant financial disincentives to pursue pediatric oncology therapeutics, including low incidence, high costs of conducting pediatric trials, and a lack of funding for early-stage research.

Observations  Review of published studies of pediatric oncology research and the cost of drug development, as well as clinical trials of pediatric oncology therapeutics at ClinicalTrials.gov, identified 77 potential drug development projects to be included in a hypothetical portfolio. The returns of this portfolio were simulated so as to compute the financial returns and risk. Simulated business strategies include combining projects at different clinical phases of development, obtaining partial funding from philanthropic grants, and obtaining government guarantees to reduce risk. The purely private-sector portfolio exhibited expected returns ranging from −24.2% to 10.2%, depending on the model variables assumed. This finding suggests significant financial disincentives for pursuing pediatric oncology therapeutics and implies that financial support from the public and philanthropic sectors is essential. Phase diversification increases the likelihood of a successful drug and yielded expected returns of −5.3% to 50.1%. Standard philanthropic grants had a marginal association with expected returns, and government guarantees had a greater association by reducing downside exposure. An assessment of a proposed venture philanthropy fund demonstrated stronger performance than the purely private-sector–funded portfolio or those with traditional amounts of philanthropic support.

Clinical Relevance  A combination of financial and business strategies has the potential to maximize expected return while eliminating some downside risk—in certain cases enabling expected returns as high as 50.1%—that can overcome current financial disincentives and accelerate the development of pediatric oncology therapeutics.

Previous estimates of drug development success rates rely on relatively small samples from databases curated by the pharmaceutical industry and are subject to potential selection biases. Using a sample of 406,038 entries of clinical trial data for over 21,143 compounds from January 1, 2000 to October 31,
2015, we estimate aggregate clinical trial success rates and durations. We also compute disaggregated estimates across several trial features including disease type, clinical phase, industry or academic sponsor, biomarker presence, lead indication status, and time. In several cases, our results differ significantly in
detail from widely cited statistics. For example, oncology has a 3.4% success rate in our sample vs. 5.1% in prior studies. However, after declining to 1.7% in 2012, this rate has improved to 2.5% and 8.3% in 2014 and 2015, respectively. In addition, trials that use biomarkers in patient-selection have higher overall
success probabilities than trials without biomarkers.

A crisis is building over the prices of new transformative therapies for cancer, hepatitis C virus infection, and rare diseases. The clinical imperative is to offer these therapies as broadly and rapidly as possible. We propose a practical way to increase drug affordability through health care loans (HCLs)—the equivalent of mortgages for large health care expenses. HCLs allow patients in both multipayer and single-payer markets to access a broader set of therapeutics, including expensive short-duration treatments that are curative. HCLs also link payment to clinical benefit and should help lower per-patient cost while incentivizing the development of transformative therapies rather than those that offer small incremental advances. Moreover, we propose the use of securitization—a well-known financial engineering method—to finance a large diversified pool of HCLs through both debt and equity. Numerical simulations suggest that securitization is viable for a wide range of economic environments and cost parameters, allowing a much broader patient population to access transformative therapies while also aligning the interests of patients, payers, and the pharmaceutical
industry.

Open-source software for "Commercializing Biomedical Research through Securitization Techniques" (with Jose-Maria Fernandez and Roger M. Stein, Nature Biotechnology 30 (2012), 964-975.