香蕉產地價格與消費地批發價格分析

臺灣果品產值占農耕產品產值的 37.2%，而香蕉產值占果品產值的 2.7%。過去香蕉產業為創匯產業，雖然近年出口量值不能與過去相比，但其種植面積為15,772.7 公頃（占果品種植面積 8.72%），仍占一席之地。相較於其他果品，青香蕉採收後由產地運輸至批發市場前必須經過催熟的程序才能成為可食用香蕉。故探討青香蕉產地價格和香蕉批發市場價格的變動，瞭解產地價格與批發市場價格將有助於對香蕉價格決定的瞭解。故本研究使用向量自我迴歸（Vector Autoregression, VAR）模型以 2010 年至 2022 年間臺灣青香蕉產地價格（以下簡稱產地價格）以及同期的批發市場價格為基礎，探討兩類市場間的價格相同性及影響決價因素。由分析結果顯示，無論是在產地或台北地區，香蕉價格皆具明顯的季節性，其中春夏季之價格較低且穩定，秋冬季價格高、到貨量少，且價格波動性大。另因產地運送至批發市場過程會產生運銷及加工成本，故批發市場價格皆較產地青香蕉高，且台北地區市場的最高平均價格發生的月份皆較產地價格最高平均價格晚一個月，可知香蕉由產地運輸至台北地區批發市場的過程可能使價格產生延遲性。而在 VAR（2）模型中，產地價格及台北地區批發價格會受到台北地區批發市場之到貨量及天然災害損失影響。由Granger 因果檢定的結果可知，產地價格、台北市批發價格，以及三重區批發價格之間應無因果影響。另衝擊反應函數結果顯示，隨著時間的推移，各價格之間的影響逐漸減弱，且台北市批發價格的影響具有延遲性，而三重區批發價格的影響力相對較小。在預測誤差變異數分解方面，產地價格對自身價格及台北地區批發價格解釋力皆為最佳，台北地區批發價格對自身價格影響相對較小。由實證結果可知，香蕉價格受台北地區批發市場到貨量及天然災害損失影響，連帶影響農民收益及民生經濟。若鼓勵農民進行產期調節、持續提升蕉農防災意識，應可有效穩定香蕉產地價格及批發市場交易價格，並保障農民收益穩定性。

關鍵字

香蕉產地價格；香蕉批發價格； VAR 模型；衝擊反應函數；預測誤差變異數分解

並列摘要

The value of fruit production in Taiwan accounts for 37.2% of the total agricultural production value, with banana production representing 2.7% of the fruit production value. While the banana industry has seen changes in export values in recent years, it still holds a significant position, with a cultivation area of 15,772.7 hectares (8.72% of the fruit cultivation area). Unlike other fruits, when deliver from the production area to the wholesale market, green bananas must undergo a ripening process before transforming to edible bananas. Therefore, exploring the variations in banana farm-gate prices and wholesale market prices is essential for understanding the price determinants of bananas.This study employed a Vector Autoregression (VAR) model as the research method, and the data are Taiwan's green banana farm-gate prices from 2010 to 2022, along with concurrent wholesale market prices, to investigate price similarities and factors influencing price determinants in the markets. The results reveal that banana prices exhibit significant seasonality, both in the production area and in the Taipei markets. Prices are lower and more stable during the spring and summer seasons, whereas they are higher in the autumn and winter seasons, characterized by lower arrivals and greater price volatility. Additionally, due to the transportation process from the production area to the wholesale market incurring delivering and processing costs, wholesale market prices are generally higher than farm-gate prices. It is also observed that the months with the highest average prices in the Taipei markets lag behind the months with the highest average farm-gate prices, suggesting a delay in pricing due to the delivering process from the production area to the Taipei wholesale markets.In the VAR (2) model, both farm-gate prices and Taipei markets wholesale prices are influenced by arrivals at the Taipei wholesale markets and losses caused by natural disasters. Granger causality test results indicate that there is no causal relationship among farm-gate prices, Taipei city market wholesale prices, and Sanchong district market wholesale prices. Furthermore, the impulse response function results show that over time, the influence between prices gradually diminishes, and Taipei city market wholesale prices exhibit delayed effects, whereas the impact of Sanchong district market wholesale prices is relatively smaller. Regarding the forecast error variance decomposition, farm-gate prices provide the best explanation for their own prices and Taipei area wholesale prices, with Taipei area wholesale prices having a relatively smaller impact on their own prices. Based on the empirical results, it is evident that banana prices are influenced by arrivals at the Taipei wholesale markets and losses resulting from natural disasters, subsequently affecting farmers' income. Therefore, encouraging farmers to engage in production period adjustments and continuously raising awareness of disaster prevention among banana farmers should effectively stabilize farm-gate prices and wholesale market prices, ultimately ensuring income stability for farmers