显示农渔民之生产系以场导向为主之型态.DOC

中英文摘要 在台灣毛豬市場是完全競爭市場，故在每一個拍賣市場中的毛豬品質是同質的。毛豬價格是以建立在拍賣交易價格來計算，因此毛豬價格的形成，完全由供需決定。毛豬產業之完全競爭、同質的特性和單一價格法則(law of one price)的前題是一致的。因此，吾人將根據檢定PPP的實證技巧，驗證毛豬價格在台灣地區是否符合單一價格法則的假說；即藉由檢定在不同地區拍賣市場之間毛豬價差的恆定性來驗證長期單一價格法則假說。如果毛豬價差是具恆定性，則長期單一價格法則有效性是成立的。 傳統的文獻是利用ADF單根檢定法來驗證購買力平價說是否成立；然眾所週知的是單一方程式之單根檢定在根接近於一時之檢力不高。為了提高檢定之檢力，吾人將採用Nybolm and Harvey (2000)提出的NH檢定法，先驗證台灣地區毛豬價格是否符合單一價格法則假說。在實證模型中，允許迴歸殘差項具同期相關性和序列相關性，並且以靴帶反覆抽樣法(bootstrap)模擬檢定統計量之臨界值。 如果有強力的證據顯示地區的毛豬價格在長期時會收斂，這是支持單一價格法則假說，亦即指出在毛豬市場上是沒有套利空間。吾人將估計毛豬價格的半衰期(half-life)及價格波動性(volatility)；並且利用(N-1)個橫斷面的收斂速度p-value(p)值﹐藉由(1-p)當作因變數和地區間之距離、價格波動、城鄉發展和運輸成本等因素當作自變數進行迴歸分析，檢視Engel and Rogers (1996,1998,1999)提出的邊境效果(border effect)的因素有那些。若單一價格法則不成立時，吾人將探討長期中毛豬市場為何存在套利空間而無法消除；並且衡量和排序地區間之距離、價格波動和城鄉發展差異等因素，根據一致性等級順序的Kendall’s係數 (Kendall’s coefficient of concordance of the rank orderings) 檢視是否因這些因素之差異，造成單一價格法則不成立。 關鍵字：單一價格法則、邊境效果、NH追蹤資料單根檢定、靴帶反覆抽樣法、半衰期 The hog market in Taiwan is a competitive market, therefore, the average quality at each auction market is relatively homogeneous. The formation of hog prices in Taiwan depends completely on the supply and demand. The features of the hog industry in Taiwan described above are consistent with all the premises of law of one price(LOP), namely the competitive market and identical goods. Hence, it is interesting to examine the hypothesis of LOP for Taiwan’s hog market. Therefore, following the empirical strategy in testing PPP, we investigate the hypothesis of long-run LOP by examining the stationarity of price differential between markets. If price differentials are stationary then the validity of long-run LOP is supported. We apply the panel data unit root tests of Nybolm and Harvey(2000) to examine the efficiency of the hog market in Taiwan by testing the hypothesis of the law of one price. In order to take into account the contemporaneous correlation and serial correlation in innovations, we apply a bootstrap to simulate our critical values. If there is strong evidence to show that regio