Risk resources and insights

Authors: Kaiyu Guan, Bin Peng, Wang Zhou (co-lead), and Zhangliang Chen

ABSTRACT

Fluctuations in temperature and precipitation are expected to increase with global climate change, with more frequent, more intense and longer-lasting extreme events, posing greater challenges for the security of global food production. A generic framework is proposed to assess the impact of climate-induced crop yield risk under both current and future scenarios by combining a stochastic model for synthetic climate generation with a well-validated statistical crop yield model. Access the research here.

Authors: Nicole D Jackson, Megan Konar, Peter Debaere, and Justin Sheffield

ABSTRACT

Global assessments of climate extremes typically do not account for the unique characteristics of individual crops. A consistent definition of the exposure of specific crops to extreme weather would enable agriculturally-relevant hazard quantification. To this end, a database is developed of both the temperature and moisture extremes facing individual crops by explicitly accounting for crop characteristics. Access the research here.

Authors: Bruce J Sherrick, Rob Myers

ABSTRACT

This paper provides insights into how cover crops and no-till practices impact agricultural risk, which can be used to strengthen the Federal Crop Insurance Program. The paper also explains how cover crops and no-till can improve farmers’ economic outcomes and support working land resilience against severe weather threats. Access the research here.

Authors: Jacopo Ponticelli, Qiping Xu, and Stefan Zeume

ABSTRACT

Using plant-level data from the U.S. Census of Manufacturer, this study looks at the short- and long-run effects of temperature on manufacturing activity. It found that while high-temperature shocks significantly increase energy costs and lower productivity for small plants, large plants are mostly unaffected. Differences in costs per unit of energy, managerial skills, and access to finance contribute to explaining why large firms are better able to adapt to climate change. Access the research here.

Authors: Eun Jeong Cha, Ph.D., Bowei Song, Yihan Jiang

ABSTRACT

Enhancing building codes is a climate-conscious way to properly manage hurricane risks. However, new building code mandates do not apply to existing buildings; these are most vulnerable and constitute the majority of building inventory. Therefore, hurricane risk mitigation efforts should be expanded to existing buildings to effectively protect residential structures. This study found that adding three structural components (accordion windows, perforated parapets, and hurricane clips) can effectively decrease average hurricane loss ratios. Access the research here

Authors: Adam Michael Bauer,  Cristian Proistosescu, Ph.D., Gernot Wagner, Ph.D.

ABSTRACT

Valuing the cost of carbon dioxide (CO2) emissions is vital in weighing viable approaches to climate policy. Conventional approaches to pricing CO2 involve evaluating trade-offs between sacrificing consumption now to abate CO2 emissions versus growing the economy, therefore enlarging one’s financial resources to pay for climate related damages later. This paper takes an alternative approach to carbon pricing by employing a financial economic model that prices CO2 emissions as an asset with negative returns. Access the research here

Authors: David Molitor, Mark Borgschulte, and Eric Yongchen Zou

ABSTRACT

Drifting wildfire smoke affects populations far from the wildfires themselves and negatively impacts the U.S. labor market. Linking satellite smoke plume data with labor market outcomes shows that an additional day of smoke exposure reduces earnings by about 0.04 percent over two years. The welfare cost of the lost earnings is on par with standard valuations of the mortality effects of wildfire smoke events. The findings suggest that labor market channels warrant greater consideration in air pollution regulations. Access the research here.

Authors: Tatyana Deryugina

ABSTRACT

Climate change is expected to have enormous effects on economic activity, yet the channels through which these materialize are not well-known.   Using high-frequency, spatially granular data, our research quantified the effect of temperature fluctuations on consumer demand, as measured by nearly 20 billion visits to a variety of business establishments in the U.S. over a 30-month period – from January 2017 through July 2019. Controlling for seasonality, secular trends, and any unobservable time-invariant location-specific characteristics, findings show that both abnormally warm and abnormally cold days increase same-day establishment visits (relative to a day where temperatures are between 0 and 3 degrees Celsius).  Access the research here.

Authors: Parker Bolstad, Sadie Frank, Eric Gesick, David Victor

ABSTRACT

Rising concerns about climate change are affecting disclosures to financial markets by looking systematically at 10-K filings from the 3000 largest U.S. publicly traded firms over the last 12 years and samplings of Official Statements from all U.S. municipal bonds. For equities, disclosure has risen sharply. Today, 60% of publicly traded firms reveal at least something about climate change, but the largest volumes of information are skewed heavily toward a few industries (e.g., electric utilities, oil & gas, mining) and concern valuation risks due to possible transition away from fossil fuels. By contrast, there is much less disclosure around the physical risks of climate change. Access the research here.

Authors: Donald J. Wuebbles

ABSTRACT

Although different analyses of climate impacts may have different national, regional, and sectoral focuses, most begin with the same premise: that future climate conditions under a set of consistent assumptions regarding human choices, as reflected in future greenhouse gas emissions, can be simulated by global climate models. Modeling of climate processes, that is, the physics, chemistry, and biology defining the Earth’s climate, are usually done at the global scale. Access the research here.

Authors: Nolan Miller, David Molitor, Eric Zou

ABSTRACT

Wildfire smoke is a major source of air pollution in the United States, resulting in serious harm to health. It travels long distances, imposing health burdens on people nationwide and making wildfire risk mitigation a national concern. The incremental mortality effects of increased air pollution are highest when the air is relatively clean. Large health benefits can likely be gained from additional improvements in air quality, even in places where air pollution levels are relatively low. Current air pollution policies that focus on the most polluted areas may benefit from incorporating these findings into their approaches. Access the research here.

Authors: Robert Trapp, Sonia Lasher-Trapp, and Kimberly Hoogewind

ABSTRACT

The effect of anthropogenically enhanced greenhouse gas concentrations on the frequency and intensity of hail depends on a range of physical processes and scales. These include the environmental support of the hail generating convective storms and the frequency of their initiation, the storm volume over which hail growth is promoted, and the depth of the lower atmosphere conducive to melting. High-resolution (convection permitting) dynamical downscaling is used to simultaneously account for these effects. Access the research here.

Authors: Robert Trapp, Sonia Lasher-Trapp, Timothy H. Raupach, Olivia Martius, John T. Allen, Michael Kunz, Susanna Mohr, Kristen L. Rasmussen, and Qinghong Zhang

ABSTRACT

Abstract | Hailstorms are dangerous and costly phenomena that are expected to change in response to a warming climate. In this Review, we summarize current knowledge of climate change effects on hailstorms. As a result of anthropogenic warming, it is generally anticipated that low- level moisture and convective instability will increase, raising hailstorm likelihood and enabling the formation of larger hailstones; the melting height will rise, enhancing hail melt and increasing the average size of surviving hailstones; and vertical wind shear will decrease overall, with limited influence on the overall hailstorm activity, owing to a predominance of other factors. Access the research here

Authors: Tatyana Deryugina and Julian Reif

ABSTRACT

Many environmental hazards have long-run health effects, but quasi-experimental studies typically measure outcomes and treatment over short time periods.  This research is about the development of a new framework for quantifying the effect of air pollution exposure on life expectancy. Using changes in wind direction as an instrument for daily sulfur dioxide levels, the research first characterizes the dynamic mortality effects of short-run exposure, then incorporates these estimates into a demographic model to quantify the lifelong effects of a permanent reduction in air pollution exposure. The study found that 90 percent of the survival benefits accrue after the first 50 years of life. Access the research here.

Authors: Ran Duchin, Janet Gao, Qiping Xu

ABSTRACT

Following damage due to an environmental incident, firms will often divest their pollutive assets. However, pollution levels do not decline after divesting because the buyers of such assets are those facing weaker environmental pressures. While the sellers may highlight their sustainable policies with the divestiture, the asset market allows firms to redraw their boundaries in a manner perceived as environmentally friendly without real improvement in pollution levels or without substantial gains. Access the research here.

Authors: Armin Tabandeh, Jung Bin Lee, Paolo Gardoni

ABSTRACT

Modeling wildfire propagation is central to wildfire risk analysis. This research features a tool for wildfire risk analysis that is both scientifically rigorous and practical for operational use. Using advanced mathematical methods and robust data-processing capabilities, the tool enables stakeholders to assess wildfire impacts on infrastructure —both before and during an event. The intent is to support proactive measures that simulate various intervention strategies that can enhance community and infrastructure preparedness. During wildfires, the research demonstrates how the tool can provide real-time predictions of fire behavior and facilitate 'what-if' scenarios to guide rapid response and containment efforts. Access the research here.

Author: Gan Zhang

ABSTRACT

This research demonstrates that hurricane risk modeling can benefit from recent breakthroughs in artificial intelligence (AI). Using an experimental system based on Google AI-physics climate model, results demonstrated a skillful seasonal prediction. Although AI does have technical limitations, integrating AI can enhance risk modeling capabilities. Access the research here.

Authors: Arkadiy Sakhartov

ABSTRACT

This study develops a formal model that demonstrates how four popular real options affect a firm’s risk. With a few qualifications, a firm’s risk is shown to be increased by real options. In addition to explaining this baseline result and the exceptions leading to risk reduction, this study develops a systematic account of how risk associated with real options derives from determinants of real options. Access the research here

Authors: Paolo Gardoni, Ph.D., Neetesh Sharma, Ph.D., Yun-Chi Yu

ABSTRACT

Because there is limited literature available to validate seismic risk assessment models, the validity and accuracy of earthquake damage predictions are often questionable. This paper evaluates the current state of practice for seismic risk assessment models, particularly those models that predict damage to buildings, transportation infrastructure, electrical power infrastructure, and potable and wastewater infrastructure based on the ground motion. This paper offers an overview of the risk assessment models’ predictive capability and suggests future research to improve these predictive models. Access the research here

Authors: Sagnik Paul, Ann C. Sychterz, Ph.D.

ABSTRACT

This paper looks at the condition of pedestrian bridges to reduce risk of degrading structures and resulting injuries to human life, cost of insurance claims, and cost of repair. Pedestrian bridges are intended to ease the movement of people while reducing the chances of accident. This paper uses machine learning to present the relation between bridge vibration, the damping coefficient with walking speed and structural age for aluminum pedestrian bridges. Access the research here

Authors: Sam Shi-Jun, Bo Li, Ph.D. 

ABSTRACT

This paper introduces a newly-developed method of estimating return levels of spatial extremes by taking spatial heterogeneity and smoothness into account. The method can help improve risk assessment and facilitate better decision-making for insuring catastrophic events. Access the research here

Authors: Paolo Gardoni, Armin Tabandeh, Neetesh Sharma, and Fabrizio Nocera

ABSTRACT

This paper focuses on three topics that represent current challenges in risk analysis of physical assets, namely, (1) the development of a systematic approach for creating digital twins of physical assets, (2) the modeling of the spatiotemporal evolution of hazards, and (3) the modeling of the time-varying vulnerability of interdependent physical assets. Accurate risk analysis requires representative mathematical models of physical assets.  These mathematical models create digital twins of reality. Access the research here.

Authors: Heitor Almeida, Viral V. Acharya, Yakov Amihud, and Ping Lui

ABSTRACT

This paper focuses on the corporate choice between financial efficiency and operational resiliency. Firms substitute between saving cash for financial hedging,  which mitigates the risk of financial default, and  spending on operational hedging,  which  mitigates  the  risk of operational default such as a failure to deliver on obligations to customers. This trade off is particularly strong for financially constrained firms and results in a positive relationship between operational spread (markup) and financial leverage or credit risk. The empirical evidence is presented that supporting  this tradeoff, the effect being pronounced for constrained firms. Access the research here.

Author: Tatyana Deryugina

ABSTRACT

Damage estimates for natural catastrophes are often imprecise. Margins of error for damage from the largest catastrophes are in the tens of billions of dollars, and geographically granular damage information is very rare. This research addressed the shortcoming by applying a computer vision model to disaster imagery data and creating a spatially detailed damage database for six major U.S. hurricanes. The resulting database—comprising nearly 1.7 million buildings—can pave the way for additional research projects, including assessments of mitigation strategies and individual-level responses to catastrophes. Access the research here.

Authors: Weichen Li, Xiaojia Shelly Zhang

ABSTRACT

The goal of this project was to create a risk-based digital twin (DT) model to enhance predictions related to the vulnerability of civil structures to climate risk. Rapidly changing environments necessitate cutting-edge solutions that are both adaptive and accurate. This project employed physics-based modeling with the advancements of stochastic optimization methods. Access the research here

Authors: Morton Lane and Jerome Kreuser

ABSTRACT

The use of optimizing models for portfolio selection and construction in the context of insurance is relatively new. Investment portfolio managers regularly rely on optimization, but underwriters are much more likely to use good old fashion trial and error, with some admittedly quite sophisticated, simulation techniques to developing underwriting portfolio strategies. The unique characteristics of insurance risk, e.g. long tails, one sided correlations etc, did not lend themselves to early optimization models but, certain technical breakthroughs have advanced optimization modeling and insurance risk is now a potentially important application. Access the research here.

Authors: Morton Lane and Jerome Kreuser

ABSTRACT

Some reinsurers use optimization procedures to generate underwriting portfolios, maximizing expected returns which are perfectly aligned with their stated risk preferences. Similar objectives apply to those who use simulation or DFA techniques. However, beyond the optimal portfolio itself, optimizers as part of their output also generate marginal economic signals, such as “implied” or “risk adjusted” probabilities which are important but underused and often misunderstood management tools. The purpose of this paper is to further illustrate the power of those economic signals.  Access the research here.