利用虛擬賽局之訂價與轉換成本投資策略研究

Abstract

本問題研究雙占廠商於單一市場上提供競爭產品並進行價格和轉換成本競爭。在有限的規劃時程內，廠商現有顧客中有部分比例會選擇在期初時進行產品升級。而當顧客選擇升級時，消費者可以從原產品供應商或其他競爭公司購買升級產品，若顧客選擇其他競爭公司進行升級，則會產生一次性的轉換成本。例如，當Android手機用戶選擇產品升級時，顧客可能會選擇繼續使用Google Android或Apple iOS智慧手機，而競爭產品間之不兼容性將產生轉換成本。為了在有限的規劃時程內最大化公司的總利潤，企業可以進行價格決策並投資以增加/減少各時期的轉換成本。我們將此問題設置為同時賽局(Simultaneous game)，結合單次抽樣虛擬賽局（Single Sample Fictitious Play, SSFP）演算法來開發近似動態規劃模型。研究顯示，SSFP演算法具優於現有算法的計算優勢，減輕了維度的詛咒。研究結果指出，轉換成本投資也許能成為現有公司的競爭障礙，以減少競爭的激烈程度。

This research studies price and switching costs competition between two firms offering competing products on a single market. Over a finite planning horizon, a portion of existing customers may choose to upgrade their products in each period. When the customers choose to upgrade, they may purchase from the original firm or the other competing firm. If the customer chooses to upgrade through the competing firm, a switching cost incurs. (e.g. When an Android smartphone user chooses to upgrade, the customer may choose to continue using Google Android or the Apple iOS smartphones and switching costs incurs due to the incompatibility between competing products.) To maximize a firm’s total profit over the finite planning horizon, firms can make price decisions and invests to increase/reduce the switching costs in each period. We formulate this problem as a simultaneous game and incorporate a single sample fictitious play (SSFP) algorithm to develop an approximate dynamic programming model. The SSFP algorithm shows the computation advantage over existing algorithms and relieves the curse of dimensionality. The results show that switching costs investments could be used competition barrier for the incumbent firm to reduce the fierceness of the competition.