再保險市場資訊不對稱之實證研究

Abstract

過去的文獻中，鮮少探討再保險市場資訊不對稱的議題。再者，資訊不對稱之理論在實證研究上，最主要被探討的議題－保額與損失機率間是否有顯著關係以及其關係之方向，文獻仍未有一致的發現與結論。本論文之目的，即為藉由探討再保險市場保額與損失機率之關係，檢視再保險市場是否存在資訊不對稱之情形；若兩者間存在顯著之關係，意即存在資訊不對稱之情形，則進一步探討其方向，藉此判定此種資訊不對稱之可能類型。 本論文使用2004年NAIC產險資料庫的資料。於整理及刪除不符邏輯之數值後，樣本中共有1,192個觀測值。為檢視再保險保額與損失機率間是否有顯著關係，先控制再保險人已知之資訊，再對殘餘之再保險比率與損失比率之條件相關性進行檢定。檢定結果未能拒絕兩者無顯著相關之假設；換言之，本論文並未發現足夠之證據，支持再保險市場之保額與損失機率有顯著關係之論點。

The purpose of this paper is to examine whether asymmetric information exists in reinsurance markets by testing the correlation between the reinsurance coverage and the loss probability of an insurer. For one thing, few papers do empirical research about asymmetric information in reinsurance markets. For another, whether insurance coverage is significantly related to loss probabilities and what the direction is if the correlation exists still don’t reach to a consistent conclusion from empirical literature. If the correlation exists; that is, asymmetric information exists, I further see the direction and try to identify possible types of this asymmetric information. The data used in this paper is from 2004 NAIC property-casualty database and there are 1,192 property and casualty insurers in the sample. After controlling reinsures’ known information, I examine the correlation by performing the conditional dependency test on the residual reinsurance ratio and the residual loss probability. However, I find no evidence for it.