電影影碟出租需求的決定因素與預測模式：以2003年至2008年上映電影為例

Determinants of the Demand for Video Rental: A Forecasting Model Approach of 2003-2008 Movies in Taiwan

黃健倫(Chien-Lun Huang)

指導教授 ： 黃俊堯

國立臺灣大學/管理學院/商學研究所/碩士(2010年)

https://doi.org/10.6342/NTU.2010.03439

摘要

本研究以台灣於2003年8月至2008年上映的電影為分析對象，藉由台北市的電影票房資料，觀察台灣電影近年變化的趨勢；並以台北板橋某間頗具規模的出租店的出租資料為例，探討近年台灣影碟出租市場的變遷，以及電影票房與電影影碟出租量之間的關連。同時藉由研究資料，分析電影票房與影碟出租集中趨勢的差異，發現電影票房的集中趨勢相當明顯，符合管理學上的80/20法則；而電影影碟的租借約80%的影碟租次集中在不到40%的電影，集中的趨勢不若電影票房來的明顯。
研究並探討影響電影票房與電影影碟出租的決定因素，從影響電影票房的迴歸模型的分析可以發現，續集電影、影片類型為動作片與奇幻片，以及在暑假、耶誕元旦與春節上映的電影的票房顯著較高，但日韓電影與其他國家（除了美國、華語國家與日韓）的電影的票房則顯著較低。另外在影響電影影碟租次的迴歸分析可以發現，電影票房對影碟出租的影響十分顯著，在控制了電影票房及其他變因後，華語電影、日韓電影與其他國家電影、電影類型為恐怖片及其他類型電影（包括音樂歌舞、傳記與紀錄片），以及在春節上映的電影的出租量顯著較低；但電影類型為動作片，以及在春節發行出租影碟的電影的出租量則顯著較高；此外，電影影碟發行時間與電影上映時間的間隔，與影碟的出租量也呈現反向的關係。
本研究進一步建立電影影碟總出租量的預測模型，經由相關測度指標顯示模型具有合理的預測能力，因此可以作為出租店在決策進片數量的一個參考指標。經由情境模擬，利用本預測模型所決策的電影影碟進片數量，可以有效減少過去出租店經營面臨的部分影碟進片太多而導致虧損，以及部分影碟進片太少讓顧客租借不到，所導致顧客滿意度下降的情形。

關鍵字

電影票房 ； 電影影碟出租 ； 集中度 ； 需求預測 ； 電影類型 ； 電影旺季

並列摘要

In this study, we use the box office in Taipei and the video rental data from a large-scale rental store in Banciao from August 2003 to December 2008 to analyze the trend of Taiwan motion industry and the relation between box office performance and video rentals. Then we analyze the differences of concentration ratio between box office and video rentals. The evidence shows that the concentration of box office is significant and conforms to the 80/20 law. However, about 80% of video rentals comes from less than 40% of the total films, which means that the concentration of video rental pattern is not so significant as movie box office.
In order to find the determinants of the film demands from box office and video rentals, we established the regression models for these two channels. According to the analysis of box office model, box office is significantly higher for sequel movies, movie genres of action movies and fantasy movies. It is also higher for those films released on summer vacation, Christmas and New Year's Day, and Chinese New Year’s Day. Conversely, the box office is significantly lower for movies from Japan, Korea and other countries (except the United States, Chinese-language countries and Japan and Korea).
On the other hand, the analysis of video rentals shows that the relation between box office and video rentals is still very significant. After controlling other variables, the video rentals are significant higher for action movie and those films released on rental market during Chinese New Year's Day. However they are significantly lower for Chinese-language movies, movies from Japan and Korea and other countries, horror movies, movies of other types (including Musical, biographies and documentaries) and movies released during Chinese New Year’s Day. We also find that duration between movie release date and video release date is inversely related to the video rental volume.
We further establish the forecasting model of the demand for video rentals. The model has reasonable predictive power, through the evidence of some measurement indicators, so it could be an indicator for the video rental shop to make the video stock decision. Through scenario simulation, the predicting model could help decide the most appropriate amount of video disk stock. It could not only effectively reduce the insatiability of customers caused by shortage of popular video disks, but also deficit resulted from overstock.

並列關鍵字

Box office ； Video rental ； Concentration ratio ； Forecasting model ； Movie genre ； Seasonality

參考文獻

翁景民、許書銘、楊君琦 (1999)，「台灣地區電影映演市場集中度分析」，新聞學研究，59：1-21。

黃詩凱 (2003)，「九０年代台灣電影市場映演市場分析：產業集中度的觀點」，傳播與管理研究，2(2)：157-174。

李怡靜 (2007)，「電影發行者的時間決策分析」，台大經濟所論文。

Ainslie, Andrew, Xavier Dreze, and Fred Zufryden (2005), “Modeling Movie Life Cycles and Market Share,” Marketing Science, 24(3), 508-517

Basuroy, Suman, Subimal Chatterjee, and S. Abraham Ravid (2003), ”How Critical Are Critical Reviews? The Box Office Effects of Film Critics, Star Power, and Budgets,” Journal of Marketing, 67(4), 103-117