Johannes Beutel

Investors’ subjective capital gains expectations are a key element explaining stock price fluctuations.  Survey measures of these expectations display excessive optimism (pessimism) at market peaks (troughs). We formally reject the hypothesis that this is compatible with rational expectations. We incorporate subjective price beliefs with such properties into a standard asset-pricing model with rational agents (internal rationality). The model gives rise to boom-bust cycles that temporarily delink stock prices from fundamentals. It quantitatively replicates many asset-pricing moments, including the observed strong positive correlation between the price dividend ratio and survey return expectations, which cannot be matched by rational expectations. 

We present a stock market model that quantitatively replicates the joint behavior of stock prices, trading volume and investor expectations. Stock prices in the model occasionally display belief-driven boom and bust cycles that delink asset prices from fundamentals and redistribute considerable amounts of wealth from less to more experienced investors. Although gains from trade arise only from subjective belief differences, introducing financial transactions taxes (FTTs) remains undesirable. While FTTs reduce the size and length of boom-bust cycles, they increase the likelihood of such cycles, thereby overall return volatility and wealth redistribution. Contingent FTTs, which are levied only above a certain price threshold, give rise to problems of equilibrium multiplicity and non-existence.

We study the effects of central bank communication about financial stability on individuals’ expectations and risk-taking. Using a randomized information experiment, we show that communication causally affects individuals’ beliefs and investment behavior, consistent with an expectations channel of financial stability communication. Individuals receiving a warning from the central bank expect a higher probability of a financial crisis and reduce their demand for risky assets. This reduction is driven by downward revisions in individuals’ expected Sharpe ratios due to lower expected returns and higher perceived downside risks. In addition, these individuals deposit a smaller fraction of their savings at riskier banks.

This paper compares the out-of-sample predictive performance of different early warning models for systemic banking crises using a sample of advanced economies covering the past 45 years. We compare a benchmark logit approach to several machine learning approaches recently proposed in the literature. We find that while machine learning methods often attain a very high in-sample fit, they are outperformed by the logit approach in recursive out-of-sample evaluations. This result is robust to the choice of performance metric, crisis definition, preference parameter, and sample length, as well as to using different sets of variables and data transformations. Thus, our paper suggests that further enhancements to machine learning early warning models are needed before they are able to offer a substantial value-added for predicting systemic banking crises. Conventional logit models appear to use the available information already fairly efficiently, and would for instance have been able to predict the 2007/2008 financial crisis out-of-sample for many countries. In line with economic intuition, these models identify credit expansions, asset price booms and external imbalances as key predictors of systemic banking crises.

This paper studies green investing in a quantitative asset pricing model with heterogeneous investors calibrated using high-quality, representative survey data of German households. We find substantial heterogeneity in green taste for both safe and risky green assets throughout the wealth distribution. Model counterfactuals show nonpecuniary benefits and hedging demands currently make green equity more expensive for firms. Yet, these taste effects are dominated by optimistic expectations about green equity returns, lowering firms' cost of green equity to a greenium of 1%. Looking ahead, we use our model to trace out the aggregate effects of information provision in an RCT and find green equity investment could potentially double when information about green finance spreads across the population. Regarding safe green assets, our model counterfactuals show that if green deposits could be offered at a 50 basis point interest rate spread, aggregate green investments in the economy could quadruple in the medium run.

We causally test alternative theories of expectation formation and asset pricing. Using a randomized information experiment we show overreaction is a key feature of individuals' belief formation. Individuals excessively extrapolate past returns and earnings growth into future returns. The average response to the price-earnings ratio is opposite to the academic consensus and individuals' reaction to stock market news depends on their perceived importance. Conditional on their beliefs, individuals' sensitivity of risky portfolio shares is consistent with the standard Merton model of portfolio choice. Our evidence suggests belief overreaction and heterogeneous subjective mental models as key ingredients to asset pricing models. 

We provide the first large-scale survey-based evidence on tail risk beliefs. Inconsistent with a static form of tail risk neglect, individuals assign large probabilities to tail events such as a financial crisis, a new pandemic, a China-Taiwan conflict, a severe energy crisis, or a Cyber attack with systemic consequences. They expect large GDP declines and substantial stock market declines, conditional on these events. However, our evidence supports dynamic tail risk neglect, as predicted by diagnostic beliefs. Extrapolation in response to news is a causal mechanism, that is able to generate pro-cyclical tail beliefs, i.e. lower perceived tail risk during booms. Moreover, individuals who view the boom state as representative, underreact to bad news. Tail risk beliefs causally affect individuals' inclination to invest into housing, equity, or precious metals. The direction of the effect depends on the type of tail event, hence tail risk is not one-dimensional, and safe haven assets are event-specific. We construct novel measures of individuals’ memory database and show they are linked to individuals’ downside tail beliefs. Our results suggest memory as a micro-foundation for diagnostic beliefs as a fruitful avenue for modeling tail beliefs and their financial implications. 

We develop a model to study when and how central banks should communicate information about asset valuations to investors, addressing the lack of guidance on optimal communication in the presence of behavioral biases. Our findings show the optimal communication strategy depends on the way investors' belief formation departs from rationality. Full disclosure is optimal only when behavioral investors' updating closely aligns with rational benchmarks. If behavioral investors overreact to positive communication, they may increase demand for risky assets excessively, causing overproduction and concentrated risk. Perhaps surprisingly, central bank communication can be undesirable when belief over- or underreaction is strongest.

We study the link between the global financial cycle and macroeconomic tail risks using quantile vector autoregressions. Contractionary shocks to financial conditions and monetary policy in the United States cause elevated downside risks to growth around the world. By tightening financial conditions globally, these shocks affect the left tail of the conditional output growth distribution more strongly than the center of the distribution. This effect is particularly pronounced for countries with less flexible exchange rate arrangements, higher foreign currency exposures, and higher levels of private sector leverage, suggesting that exchange rate policies and macroprudential policies can mitigate downside risks to growth.

We use the Covid-19 pandemic as a natural experiment to estimate the effects of a global disaster shock on economic activity, international trade, and financial markets. To identify the shock, we exploit cross-country variation in the timing and intensity of pandemic-induced social distancing at daily frequency. An unexpected reduction in human mobility relative to pre-pandemic levels leads to significant declines in high-frequency measures of global economic activity, such as daily nitrogen dioxide concentrations and daily maritime trade carried by large cargo ships. Global stock markets decline, and sovereign credit spreads persistently widen after the pandemic disaster shock.