農業是民眾生活所必須依賴的基礎產業之一。然而台灣地域範圍較小，農業產區集中，農產品種類繁多，農業生產因各類自然災害風險所產生損失呈現較明顯的正相關性，風險不易分散；另一方面，農業保險因信息不對稱和勘損、監督難度大，逆選擇和道德風險問題也比較嚴重。因此在台灣開辦農業保險並不容易。
2015年台灣推出了第一款高接梨保險，之後一年裡又增加了芒果保險。但是由於保費較高，前兩年投保率很低，如芒果保險在第一年只賣出6張保單。面對開辦農業保險所面臨的困難，本文針對台灣的農業保險及其系統性風險進行研究，提出了兩方面可能的風險分散策略：擴大地理區域合作以分散風險和增加承保農作物品種來分散風險。研究通過建立地理空間相關性模型、系統性風險保費附加模型以及風險分散效果評價模型，對台灣實證數據分析研究發現：
（1）當水稻、小麥作物產區地理距離大於900公里時，產量狀況呈現不相關。
（2）通過擴大地理區域合作，可以有效降低台灣水稻作物的系統性風險問題，但是對玉米作物效果並不明顯。這是由於台灣當地不同縣區玉米生產本身風險正相關性較弱，也說明不同種類的農作物所面對的系統性風險狀況有所不同。
（3）在承保農作物品種較少時，增加承保種類，可起到明顯的風險分散效果；當承保台灣主要的十種農作物以後，繼續增加承保品種，風險分散效果提升不明顯。

Agriculture is one of the basic industries and important to common people. Be-cause of the small area of Taiwan, the agriculture is very concentrated. So the risks among counties are obviously positive correlated and difficult to hedge. Meanwhile, moral hazard and adverse selection are also big problems due to information asymmetry and high cost of damage prospecting and farmers’ behavior surveillance. As a result, operating agricultural insurance is not easy in Taiwan.
In 2015 and 2016, insurance of pears and mangoes began to be sold respectively. However, the prices are relatively high and little farmers bought them. In fact, only 6 policies of mango insurance are sold in 2016. So this article aims at studying agricultur-al insurance and the systemic risk in Taiwan. Pooling some counties of China together with Taiwan and increasing underwriting crop species are two possible ways to diver-sify the systemic risk. The results are below:
(1) When the distance between two counties is no less than 900km, the yields of rice or corn are unrelated.
(2) Pooling some Counties of China do help the diversification of systemic risk for rice, but it’s useless for corn. The systemic risks for different species are not identical.
(3) When the insurance covers few species, increasing more species can help di-versify the systemic risk. But if the 10 main species are already covered, add-ing more underwriting species do not help.

Bühlmann, H., 1970, Mathematical Methods in Risk Theory (New York: Spring-er-Verlag).
Cummins, J. D., & Trainar, P. (2009). Securitization, Insurance and Reinsurance. Jour-nal of Risk and Insurance, 76, 463-492.
Diaz-Caneja, M. B., Conte, C. G., Pinilla, F. J. G., Stroblmair, J., Catenaro, R., & Dittmann, C. (2009). Risk Management and Agricultural Insurance Schemes in Europe, JRC Reference Report, EU-23943, EN-2009, European Commission.
Guyon, X. (1995). Random Fields on a Network: Modeling, Statistics and Applications (New York: Springer-Verlag).
Hellmuth, M. E., Osgood, D.E., Hess, U., Moorhead, A., & Bhojwani, H. (2009). In-dex Insurance and Climate Risk: Prospects for Development and Disaster Man-agement. Climate and Society, 2, International Research Institute for Climate and Society (IRI), Columbia University, New York.
Knight, T. O., & Coble, K. H. (1997). Survey of U.S. Multiple Peril Crop Insurance Literature Since 1980. Review of Agricultural Economics. 19, 128-156.
Kramer, R.A.(1983) Federal Crop Insurance 1938-1982. Agr. History 57(April), 181-200.
Mario, J. M., & Joseph, W. G. (1997). Systemic Risk, Reinsurance, and the Failure of Crop Insurance Markets. Amer. J. Agr. Econ. 79(February), 206-215.
Mehr, R. I., Cammack, E., & Rose, T. (1985). Principles of Insurance, 8th ed.
Miranda, M., & Glauber, J. (1997). Systemic Risk, Reinsurance and the Failure of Crop Insurance Markets. American Journal of Agricultural Economics, 79, 206-215.
Okhrin, O., Odening, M., & Xu, W. (2013). Systemic Weather Risk and Crop Insur-ance: the Case of China. Journal of Risk and Insurance, 80(2), 351-372.
Wang, H. H., & Zhang H. (2003). On the Possibility of a Private Crop Insurance Mar-ket: A Spatial Statistics Approach. The Journal of Risk and Insurance, 70(1), 111-124.
楊明憲，2011，國外實施各類型農業保險經驗之探究，行政院農業委員會網站，http://www.coa.gov.tw/view.php?catid=59。
中國農業保險發展研究報告，2012，中國農業出版社。