本研究分析汽車失竊風險的風險管理問題、及台灣汽車竊盜險市場上的訊息不對稱 問題。 管理失竊風險的方法，包括：購買保險、或進行自我防衛。本研究的第一個主題， 探討個人究竟將市場保險、與自我防衛，視為替代、還是互補？在道德風險的角度下， 兩者可能是相互替代的；但Ehrlich 與Becker (1972)指出，當保險公司觀察被保險人的 自我防衛行為，且當成保費減費依據時，自我防衛與市場保險，又可能是相互互補的。 過去，由於資料的限制，少有實證驗證這個問題。而本研究由於取得保險公司觀察 汽車加裝衛星定位防盜設備的自我防衛行為、進行保費減費前、後的獨特的資料，可以 比較考量自我防衛，進行保費減費前、後，自我防衛與市場保險間，互補、或替代關係 的變化。 其次，在保險市場訊息不對稱的議題上，由於保險市場存在的隱藏訊息，可能造成 訊息不對稱，讓保險公司虧損、或市場萎縮。因此，本研究的第二個主題，就是找出汽 車竊盜險市場上，重要的隱藏訊息，也就是被保險車輛的停放地點，是否位於犯罪熱點， 並檢測這項隱藏訊息，能否顯著解釋汽車失竊機率、及保險契約選擇機率，並進一步檢 定控制這個隱藏訊息後，是否可以減緩訊息不對稱的嚴重程度。 近年的文獻研究也指出，保險市場可能有多維度的訊息不對稱問題存在。因此，本 研究的第三個主題，旨在發掘台灣汽車竊盜險市場上，是否有風險型態、與風險趨避兩 個不同維度的訊息不對稱問題，形成逆選擇、與有利選擇同時存在的現象。在實證上， 透過Cohen 與Einav (2007)的方式，估計出個人絕對風險趨避係數，並利用這項隱藏訊 息，進行多維度訊息不對稱問題的檢證。 This research investigates the risk management strategies on the car theft risk, and the asymmetric information problems on Taiwan’s car theft insurance market. The risk management strategies on car theft risk include purchasing car theft insurance, and self protection. The first topic of this research is to investigate that whether market insurance and self protection are complementary or substitutive? From the view point of moral hazard, market insurance and self protection are substitutive. But, Ehrlich and Becker (1972) pointed out that market insurance and self protection could be complementary while the insurance company can observe the self protection activities of the insured and give some premium discount accordingly. Few empirical literatures have ever studied on this issue by the constraint of available data. This research collects unique data of market insurance and self protection about the car theft risk, before and after the insurance company observed the self protection of burglar alarm and gave premium discount accordingly. It allows us to investigate the relationship between market insurance and self protection of car theft before and after the observation change of the insurance company. On the other hand, in the issue of asymmetric information, there could be loss for the insurance company, and the insurance market could be faded because of some hidden information existed. The second topic of this research is to find the important hidden information of the car theft insurance market. Whether the insured car is parked in crime hot spot or not is the hidden information we try to test. Our key point is to test its explanatory ability on the probability of car theft accident, and on the probability of purchasing high coverage car theft insurance contract. Furthermore, we test that whether the asymmetric information phenomenon is mitigated through this hidden information is controlled. The literatures in recent years also referred that there could be multi-dimensional asymmetric information existed. Hence, the third issue of this research aims at that whether there is multi-dimensional asymmetric information existed, such as: risk type as well as risk aversion, and which makes the coexistence of adverse selection and advantageous selection. We estimate the absolute risk aversion coefficient by the method of Cohen and Einav (2007), and we test the multi-dimensional asymmetric information through this hidden information.
