Essays on Risk and Uncertainty: Insights from Behavioral Economics
MetadataShow full item record
Essays on Risk and Uncertainty: Insights from Behavioral Economics Sensitization, Excess Volatility, and Extraordinary Persistence It is well-documented that stock prices are more volatile than their underlying fundamentals. A consensus has emerged that time-varying risk-premia are the likely source of this excess volatility, but no consensus has emerged regarding the source of the time-varying risk-premia. Recent microeconomic research suggests that one likely source is that risk preferences are time-varying. This same literature also suggests that variation in risk preferences can be extraordinarily persistent, on the order of decades (see Malmendier, Tate, and Yan (2011)); however this persistence has not been explained by conventional models. In this paper, we derive a model to explain both excess volatility and extraordinary persistence. To do so, we draw from the literatures of medicine, psychology, and behavioral economics. Our basic framework is that people have adaptive emotions and that these adaptive emotions create adaptive risk-aversion. This process is called sensitization, which implies that people become more risk-averse after negative shocks (Kandel 2000). To conduct our analysis, we construct an overlapping generations model of the macroeconomy to study the effect of allowing agents to be sensitized to risk. We find two main results. First, the adaptive nature of risk preferences combined with the finite horizons of agents imply that economic activities, such as investment, are too risky on the intensive margin. Second, excess risk-intensity combined with the availability heuristic implies that agents undertake too little risk (too little investment) on the extensive margin. In order to characterize the optimal monetary policy, we follow Tirole (2006), who models risk through liquidity shocks, and we derive three policy implications for policymakers. First, diversification blunts the impact of time-varying risk aversion. As a result, there is a reason to think that equity financing, under which risk diversification is easier to achieve, leads to fewer risk distortions and faster steady-state growth. Second, countercyclical risk-aversion favors countercyclical monetary policy. Third, short-term asset purchases are shown to exacerbate risk distortions. In our model, monetary policy results in greater stabilization and faster growth when conducted through long-term asset purchases such as Quantitative Easing and Operation Twist. Reference Points, Leaders, and Organizational Culture The work of Akerlof and Kranton (2005) suggests that an organization's culture affects individual behavior by shaping preferences. Yet, within the economics literature, little is known regarding the properties of the optimal culture. In this paper, we use an agency setting to determine the cultural properties that best foster incentives. To do so, we break culture down into three components: a type of performance metric (either production or cost), an expected performance level or target (that serves as the reference point, following Koszegi and Rabin (2006, 2007)), and the degree to which an agent's effort influences the benchmark (referred to as acclimation by Koszegi and Rabin (2006, 2007)). Properties of culture affect agents' consideration of effort. Under the reference-dependent preferences of Koszegi and Rabin (2006, 2007), higher effort increases the likelihood of beating the agent's target (or reference point) as well as increasing the agent's reference point. The magnitudes of these two effects depend critically on the degree of acclimation, whereas the signs of these effects depend on the type of metric used. We present three general findings. First, organizations that rely on production metrics have incentives at least as strong as those relying on cost metrics. Second, the impact of acclimation depends critically on the type of metric used. Under cost metrics, higher acclimation leads to stronger incentives. Under production metrics, higher acclimation leads to weaker incentives. Third, the optimal culture is characterized by production metrics and unacclimating reference points, which we show have implications regarding organizational tenure policies. We conclude with a discussion of testable implications. We refer to the psychology literature and argue that production metrics are most likely to emerge when production is characterized by a high degree of uncertainty, such as in sales. Our model's main prediction is that in these types of environments, we would expect to have rapid production and low tenure in order to lower acclimation. In contrast, environments in which costs are more uncertain are more likely to have cost metrics, which favor longer tenure and loose deadlines in order to generate more acclimating reference points. The Precautionary Principle in Product Markets There any many differences between the U.S. and European regulation, but one notable difference concerns assessments of risk. U.S. and European regulation are concerned about different sources of risk and these sources of risk do not always overlap. As noted by Vogel (2003), U.S. regulation gives more consideration to risk concerning environmental harms, carcinogens in food, and endangered species, whereas European regulation emphasizes risks inherent in biotechnology and carbon emissions. In fact, to justify the regulation of biotechnology, Europeans give explicit emphasis to the Precautionary Principle: faced with an irreversible choice, it is better to presume significant harm. However, when it comes to carcinogens in food, Europeans are relatively more willing to bear the risks. In this paper, we use an agency setting to determine how regulators should manage the risks inherent in new products while not placing an undue burden on potential innovators. Faced with a product quality, they can adhere to the Precautionary Principle and presume harm. Alternatively, they can adhere to the Presumption of Innocence and presume the product is harmless. This paper analyzes which is better. There are two assumptions that separate our analysis from the literature. First, we consider a static framework in which no new information arises. Second, we assume that the equilibrium risk is endogenous. Entrepreneurs' can mitigate harm if given the appropriate incentives and their choices to mitigate harm will be influenced by the regulatory framework chosen by the regulators. We present three main findings. Under the extreme assumptions of risk neutral entrepreneurs, an absence of limited liability constraints, and low levels of potential harm, we show that either the Precautionary Principle or a Presumption of Innocence can achieve the first best outcome when faced with a product of uncertain quality. However, under less extreme assumptions, we identify two factors that favor an approach more consistent with the Precautionary Principle. First, if the dominant concern of the regulators is the limited liability constraint, then relying on the Precautionary Principle will best extract rent. Second, under larger levels of harm, the introduction of agency costs (either due to risk aversion or limited liability) will interact with dynamic complementarities. As the cost to incentivize risk mitigation increases, the equilibrium likelihood of severe harm will rise, and the principle will be more likely to prevent the product from coming to the market. Preventing the product from entering the market reduces the incentives to mitigate harm further. In our model, this dynamic complementarity can only exist when the potential harm is large enough that the product's net benefit to society may be negative.
- Economics