Die Minderung von Datenknappheit bei der Bewertung von Ernteversicherungen anhand räumlicher Informationen: Ein Propagation-Separations-Ansatz
Zusammenfassung der Projektergebnisse
We extended the Propagation-Separation approach in the temporal dimension and developed an adaptive local parametric model for estimating crop yield. Our empirical results demonstrate that the proposed local parametric approach selects reasonable intervals of parameter homogeneity and helps mitigate the non-stationarity of crop yield. A rating analysis suggests that the proposed model shows potential to obtain more accurate rates than commonly used methodology. Investigating the poolability of spatial information, we applied the Propagation-Separation approach on estimating the location values. Typically “Bodenrichtwerte” that reflect the average location value of land plots within a specific area are used. These values constitute an important source of information that contributes to price transparency on land markets and are provided by publicly appointed expert groups (Gutachterausschüsse). We find based on empirical investigations, these values tend to underestimate location values. As an alternative, we suggest advanced statistical tools (Propagation-Separation Approach). The complaints over spatial disparities between insurance payments in adjoining counties has drawn considerable attention. Given that the disparities in payments mainly result from the disparities in county yields, the usage of alternative data sources of crop yields has been proposed as a potential solution. Our results showed that neither of two data sources (NASS and RMA) can systemically ‘fix’ the payment disparity among adjacent counties since they are just two county yield estimates based on sampled data after all. If the objective were only to mitigate the spatial disparity and smooth the indemnity payments among farmers, statistical smoothing technique across counties yields show potential. However, it should be noted that it also may increases the basis risk between actual crop losses and indemnity payments. Finally, in the longer term, an alternative approach is to utilize the increasing capability to map yields and yield risk spatially on a fine (sub-county) grid. This may open the door to discarding politically defined county boundaries to an alternative where ‘area’ is defined by homogeneous agronomic regions. This could increase the correlation of the ‘area’ with the individual producer yield and improve area designs as a risk management tool. https://www.hu-berlin.de/de/pr/nachrichten/august-2018/nr_180810_00
Projektbezogene Publikationen (Auswahl)
-
(2018): Adaptive local parametric estimation of crop yields: implication for crop insurance ratemaking. European Review of Agricultural Economics, 45(2): 173-203
Shen, Z., Odening, M., Okhrin, O.
-
(2018): Estimating location values of agricultural land. German Journal of Agricultural Economics, 66(3): 188–201
Helbing, G., Shen, Z., Odening, M., Ritter, M.
