ANALYSIS AND MODELING OF FIRM LEVEL RISK AND PRODUCER BEHAVIOR IN AGRICULTURE WITH AN EMPHASIS ON PRODUCTION RISK, INSURANCE USE, AND ACTUARIAL SYSTEMS ANALYSIS

• Sponsoring Institution: National Institute of Food and Agriculture
• Project Status: COMPLETE
• Funding Source: HATCH
• Reporting Frequency: Annual
• Accession No.: 0222976
• Project Start Date: Jul 6, 2010
• Project End Date: Jul 5, 2015
• Program Code: [(N/A)]- (N/A)

Recipient Organization
TEXAS A&M UNIVERSITY
750 AGRONOMY RD STE 2701
COLLEGE STATION,TX 77843-0001

Performing Department
Agri Economics

Non Technical Summary
The recent financial collapse of 2008-2009 and the unprecedented volatility in agricultural and financial prices alike indicate that the global marketplace has entered into a new era of risk. Moreover, the prospect of climate change and forecasts of increasingly tight resource demands around the globe highlight the urgency of these and other risks facing society. Understanding how firms should and do interact with risks stemming from changing and uncertain institutional structures, marketplaces, and natural environments has thus arguably never been more important. Yet, many of these risks and their implications for firms are still not well understood in application. A firm's primary profit and solvency risk mainly stems from output price and production/cost volatility, dependencies between them, as well as uncertainties surrounding changing institutional structures. Accurate modeling of these elements as well as the tools that firms use to mitigate these risks are important for identifying and formulating strategies to deal with them. While several important theoretical advances have been made in the risk modeling literature in recent decades, a sound empirical understanding of these advances is still lacking, as are many of the technical and operational links that are needed to implement such advances in real-world applications. Furthermore, historically much of the empirical work in modeling firm risk has been conducted in highly stylized frameworks that are not consistent with the dynamics of the real-world. This fact is quite troubling as it impacts the generality and accuracy of much of the existing work on firm risk modeling. Addressing these gaps is the general objective of this research project. More specifically, in applied field such as agricultural economics much attention has been focused on investigating models for describing risks facing the firm, searching for the "best" models, as well as on designing and investigating desirable frameworks within which to do so (see e.g., Just and Weninger, 1999; Turvey and Zhao, 1999; Ker and Goodwin, 2000; Wang and Zhang, 2002; Atwood et al., 2003; Ramirez Misra, and Nelson, 2003). Yet, much disagreement remains about what the best approaches for modeling firm risks are in different situations, what criteria should be used, and even what frameworks are most appropriate. As it regards the modeling of production risk, traditionally efforts in this domain have focused on identifying suitable candidate models in in-sapmle frameworks (Day, 1965; Gallagher, 1986; Nelson and Preckel 1989; Nelson, 1990; Ramirez, 1997; Goodwin and Ker, 1998; Just and Weninger, 1999; Ker and Goodwin, 2000; Ramirez Misra, and Nelson, 2003; Ker and Coble, 2003; Sherrick et al., 2004) but more recently attention has begun to focus more on (distributional) forecasting performance and associated frameworks (Norwood, Roberts, and Lusk, 2004; Lanoue et al., 2010; Woodard et al., 2010; Woodard, 2010). Accordingly, this project seeks to gain a better understanding of these impacts in order to design more effective and efficient frameworks and methods for modeling the risks facing agribusiness firms.

Animal Health Component

34%

Research Effort Categories

Basic

33%

Applied

34%

Developmental

33%

Classification

Knowledge Area (KA)	Subject of Investigation (SOI)	Field of Science (FOS)	Percent
601	0199	3010	20%
601	0420	3010	10%
601	0420	3100	10%
601	0430	3010	10%
601	0430	3100	10%
601	1510	3010	10%
601	1520	3010	10%
601	1820	3010	10%
601	1820	3100	10%

Knowledge Area
601 - Economics of Agricultural Production and Farm Management;

Subject Of Investigation
0420 - Weather; 1510 - Corn; 0430 - Climate; 1520 - Grain sorghum; 1820 - Soybean; 0199 - Soil and land, general;

Field Of Science
3010 - Economics; 3100 - Management;

Keywords

production risk, agricultural economics, insurance, spatial

econometrics, risk management

Goals / Objectives
Objectives 1. Analyze the nature of risks facing modern agribusiness firms and issues confounding the modeling of such risks. 2. Investigate the interaction between insureds and insurers, along with the impact of the insurance complex on agricultural producers. 3. Develop and modify frameworks for modeling firm level risk, with an emphasis on copulas, spatial analysis, weather, and climate impacts. 4. Investigate the impact of data constraints on risk estimation in applied setting, and employ cross-validated estimation techniques (where necessary) in order to improve estimation and inform optimal decision-making under risk. 5. Explore the impacts of risk and uncertainty on strategic and risk management decisions by producers, managers, and agribusiness firms.

Project Methods
The objectives of this research project will be fulfilled through the development and application of frameworks and methodologies for improving firm risk analysis using statistical, economic, and mathematical tools, primarily in simulation and econometric frameworks. Motivations for adoption of research approaches will be driven by knowledge of relevant institutional structures, recent theoretical and methodological developments, real-world implementation considerations, and stakeholder needs. Research problems will be investigated using accepted methods, and results disseminated via scholarly publications and professional presentations. More specifically, the following approaches will be pursued and employed: 1. Investigate and develop frameworks for assessing sampling variability in the context of multivariate risks, including applications to crop yield distribution estimation, copula modeling, and assessment of productivity gains using cross-validation and machine learning techniques. For example, overfitting problems in crop yield distribution estimation can have severe impacts the resulting estimates of insurance rates and optimal behavior. Thus, models will be developed to explain the impact of small sample weather variability on risk estimation for firms and insurance systems, particularly in estimating insurance risk, by utilizing state-of-the art spatial econometric and actuarial techniques. This work will involve research and programming in the realm of spatial econometrics, for example, in the development of spatial panel Tobit models necessary to model typical insurance loss data. 2. Develop models to optimally combine data from various sources in order to optimize credibility of resulting estimators utilizing classical and Bayesian statistical frameworks. Credibility is a concept from the actuarial literature which seeks to quantify the relative information content and consistency of such content in any particular set of data when assessing risk. Credibility theory will be applied to problems in applied agricultural risk analysis in order to improve estimation efficiency. 3. Model impacts of risk perceptions on risk management decisions, including insurance purchases and strategic investments, via utilization of survey and microeconometric techniques. Gaining an understanding of producer responses to real and perceived risk is at both the foundation and horizon of economic and financial disciplines. Projects under this approach will be pursued to investigate, for example, the impacts of information and perceptions of climate change on strategic investments and subjective asset valuation using a blend of psychometric and econometric risk elicitation techniques.

Progress 07/06/10 to 07/05/15

Outputs
OUTPUTS: Research as outlined in the proposal was conducted and presented at several meetings and conferences, and also resulted in several academic manuscripts, which are either in the review or acceptance stage currently. Work was also conducted with graduate students, resulting in conference presentations, theses, and academic papers which will be submitted to academic journals for publication. PARTICIPANTS: Nothing significant to report during this reporting period. TARGET AUDIENCES: Nothing significant to report during this reporting period. PROJECT MODIFICATIONS: PI has transferred to Cornell University.

Impacts
The results from the work conducted lead to a better understanding of risks facing firms and issues confounding the modeling of such risks. The knowledge gained from this research helps inform policy decisions as well as benefits producers by allowing for a better understanding of the risks they face. Continuing on previous and concurrent work, methods were developed to more accurately model conditional distributions for farm production which allow for improved estimation of insurance risk, weather/climate risk, and also allowed us to address and answer some long standing questions regarding the impact of biotechnology on risk assessment in the Federal Crop Insurance Program. Mixture modeling methods were also developed and published which allow for more efficient estimation of farm and county level production estimates. Work was also conducted on actuarial systems and methods used in the Federal Crop Insurance program that allow for a better understanding of the functioning of that program. Copula methods were also investigated for modeling farm and county level risk, and their impacts on insurance and risk analysis assessed, the results of which lead to a better understanding of the modeling of mutlidimensional risks which are important for informing policy decisions in the Federal Crop Insurance Program, and also improve producer understanding of the risks they face from certain types of risk mitigation activities including insurance purchases. Several other risk related issues were also investigated, with an emphasis on how model or market risk impacts producers and systems affecting producers. Last, work on skip row corn and implications for the insurance program were developed and published.

Publications

• Ghosh, S., J.D. Woodard, D. Vedenov, Efficient Estimation of Copula Mixture Models: An Application to the Rating of Crop Revenue Insurance, Agricultural and Applied Economics Association Selected Session Paper, Pittsburgh, PA (July, 2011).
• Alarcon, E., J.D. Woodard, E. Rister and D. Leatham, An Analysis of Investments in Apartments: Is There Value in a Portfolio Approach Intellectbase Academic Conference San Antonio, TX (March 25-26, 2011).
• Woodard, J.D., B.J. Sherrick, G.D. Schnitkey, Actuarial Impacts of Loss Cost Ratio Ratemaking in U.S. Crop Insurance Programs, Journal of Agricultural and Resource Economics, 36-1 (April 2011):211-228.
• Woodard, J.D., K. Ward, G.D. Schnitkey, P. Burgener, and A.D. Pavlista, Government Insurance Program Design, Incentive Effects, and Technology Adoption: The Case of Skip-Row Crop Insurance, American Journal of Agricultural Economics (2012).
• Leidner, A.J., R.D. Lacewell, M.E. Rister, J.D. Woodard, A.W. Sturdivant, and J. White, Seawater Desalination for Municipal Water Production, Desalination (2012).
• Woodard, J.D., B.J. Sherrick, Estimation of Mixture Models using Cross-Validation Optimization: Implications for Crop Yield Distribution Modeling, American Journal of Agricultural Economics 93-4(2011): 968-982.
• Woodard, J.D., N. Paulson, D. Vedenov, and G. Power, Efficiency in the Modeling of Dependence Structures: An Application of Alternative Copulas to Agricultural Insurance Rating, Agricultural Economics 42-IS1(Nov. 2011): 101-112.
• Woodard, J.D., T.M. Egelkraut, P. Garcia, and J.M.E. Pennings, Effects of Full Collateralization in Commodity Futures Investments, Journal of Derivatives and Hedge Funds, 16- 4 (2011): 472-488
• Woodard, J.D., K. Ward, G.D. Schnitkey, P. Burgener, and A.D. Pavlista, Economics of Skip-Row Corn in the Central Great Plains: Production and Insurance Implications, Society for Risk Analysis 2011 Annual Meeting, Charleston, SC (December, 2011).
• Woodard, J.D., A Conditional Distribution Approach to Assessing the Impact of Weather Sample Heterogeneity on Yield Risk Estimation and Crop Insurance Ratemaking, Agricultural and Applied Economics Association Selected Session Paper, Pittsburgh, PA (July, 2011).
• Woodard, J.D., B.J. Sherrick, Estimation of Mixture Models using Cross-Validation Optimization: Implications for Crop Yield Distribution Modeling, Annual Meeting of the SCC-76 Economics and Management of Risk in Agriculture and Natural Resources, Atlanta, GA (March, 2011).
• Leidner, A.J., R.D. Lacewell, M.E. Rister, J.D. Woodard, A.W. Sturdivant, and J. White, Seawater Desalination for Municipal Water Production, Selected Paper prepared for presentation at the Southern Agricultural Economics Association Annual Meeting, Corpus Christi, TX, February 5-8, 2011
• Adusumilli, N.C., R.D. Lacewell, M.E. Rister, J.D. Woodard, A.W. Sturdivant, Effect of Agricultural Activity on River Water Quality: A Case Study for the Lower Colorado River Basin, Selected Paper prepared for presentation at the Southern Agricultural Economics Association Annual Meeting, Corpus Christi, TX, February 5-8, 2011.
• Li, Yariu, J.D. Woodard, and D. Leatham, The Causality of Foreign Direct Investment and Its Effects on Economic Growth: Re-estimated by a Directed Graph Approach, Selected Paper prepared for presentation at the Southern Agricultural Economics Association Annual Meeting, Corpus Christi, TX, February 5-8, 2011

Progress 01/01/10 to 12/31/10

Outputs
OUTPUTS: Research as outlined in the proposal was conducted and presented at several meetings and conferences (regional, national, and international), and also resulted in several academic manuscripts, which are either in the review or acceptance stage currently. Work was also conducted with graduate students, resulting in conference presentations and academic papers which will be submitted to academic journals for publication. Publications are reported in another section of this report. PARTICIPANTS: Nothing significant to report during this reporting period. TARGET AUDIENCES: Nothing significant to report during this reporting period. PROJECT MODIFICATIONS: Nothing significant to report during this reporting period.

Impacts
The results from the work conducted lead to a better understanding of risks facing firms and issues confounding the modeling of such risks. The knowledge gained from this research helps inform policy decisions as well as benefit producers by allowing for a better understanding of the risks they face. First, methods were developed to more accurately model conditional distributions for farm production which allow for improved estimation of insurance risk, weather/climate risk, and also allowed us to address and answer some long standing questions regarding the impact of biotechnology on risk assessment in the Federal Crop Insurance Program. Mixture modeling methods were also developed which allow for more efficient estimation of farm and county level production estimates. Work was also conducted on actuarial systems and methods used in the Federal Crop Insurance program that allow for a better understanding of the functioning of that program. Copula methods were also investigated for modeling farm and county level risk, and their impacts on insurance and risk analysis assessed, the results of which lead to a better understanding of the modeling of mutlidimensional risks which are important for informing policy decisions in the Federal Crop Insurance Program, and also improve producer understanding of the risks they face from certain types of risk mitigation activities including insurance purchases. Several other risk related issues were also investigated, with an emphasis on how model or market risk impacts producers and systems affecting producers.

Publications

• Li, Yariu, J.D. Woodard, and D. Leatham, "The Causality of Foreign Direct Investment and Its Effects on Economic Growth: Re-estimated by a Directed Graph Approach," Selected Paper prepared for presentation at the Southern Agricultural Economics Association Annual Meeting, Corpus Christi, TX, February 5-8, 2011
• Woodard, J.D., N. Paulson, D. Vedenov, and G. Power, "Efficiency in the Modeling of Dependence Structures: An Application of Alternative Copulas to Agricultural Insurance Rating," Risk Analysis 2010, the Seventh International Conference on Computer Simulation in Risk Analysis and Hazard Mitigation, Algarve, Portugal (September, 2010).
• Woodard, S., N. Paulson, J.D. Woodard, and K. Baylis, "A Spatial Analysis of Agricultural Cash Rents," Agricultural and Applied Economics Association Selected Session Paper (July, 2010).
• C. Lanoue, B.J. Sherrick, G.D. Schnitkey, N.D. Paulson, and J.D. Woodard, "Efficiency Evaluation of Alternative Yield Models for Crop Insurance Rating Applications," Agricultural and Applied Economics Association Selected Session Paper (July, 2010).
• Woodard, J.D., and Nicholas D. Paulson, "Adverse Selection in Crop Insurance Markets: Impacts of Loss Expectations on Insurance Demand," Annual Meeting of the SCC-76 "Economics and Management of Risk in Agriculture and Natural Resources," Destin, Florida (March, 2010).
• Woodard, J.D., B.J. Sherrick, G.D. Schnitkey, "Crop Insurance Ratemaking under Trending Liabilities," Journal of Agricultural and Resource Economics (2011).
• Woodard, J.D., G.D. Schnitkey, B.J. Sherrick, N. Lozano-Gracia, and L. Anselin, "A Spatial Econometric Analysis of Loss Experience in the U.S. Crop Insurance Program," Journal of Risk and Insurance (2010).
• Woodard, J.D., B.J. Sherrick, and G.D. Schnitkey, "Revenue Risk Reduction Impacts of Crop Insurance in a Multi-Crop Framework," Applied Economic Perspectives and Policies (Formerly the Review of Agricultural Economics; 2010).
• Woodard, J.D., T.M. Egelkraut, P. Garcia, and J.M.E. Pennings, "Effects of Full Collateralization in Commodity Futures Investments," Journal of Derivatives and Hedge Funds, Volume 16, Issue 4 (February 2011).
• Woodard, J.D., B.J. Sherrick, "Estimation of Mixture Models using Cross-Validation Optimization: Implications for Crop Yield Distribution Modeling," Annual Meeting of the SCC-76 "Economics and Management of Risk in Agriculture and Natural Resources," Atlanta, GA (March, 2011).
• Leidner, A.J., R.D. Lacewell, M.E. Rister, J.D. Woodard, A.W. Sturdivant, and J. White, "Seawater Desalination for Municipal Water Production," Selected Paper prepared for presentation at the Southern Agricultural Economics Association Annual Meeting, Corpus Christi, TX, February 5-8, 2011