BOT計畫權利金談判模式之研究

Abstract

按促參法與相關子法規定，在BOT計畫之特許契約內需載明權利金及費用負擔事項，從法規、工程會研究報告及所有權概念，政府收取權利金應符合特許權、所有權及經營權之資源使用付費概念。政府所估計之權利金僅能參照其他國家相關BOT計畫案例及BOT計畫特性，無法判斷民間所提送權利金之合理範圍，故本研究希望建立BOT計畫權利金談判模式，提供政府與最優申請人在權利金議題談判之參考。 本研究所建構之權利金談判模式除計算合理之權利金外，更期望透過政府與最優申請人在談判過程的互動，剖析BOT計畫中政府與最優申請人之談判行為，供政府與最優申請人作為決策之參考，並改善以往兩造進行權利金談判曠廢日時之窘境。本研究利用二階規劃方法研擬政府與最優申請人之談判模式，政府追求本身財務回收率最大化，最優申請人之目標為獲利能力最大，並設定權利金為兩階層之決策變數，分別建立三種不同收取方式之權利金談判模式（1）固定式（2）營收比例（3）運量比例，並以MATLAB撰寫啟發式求解法，設計多次談判之求解流程，直至求得妥協解為止，另外，本研究將談判次數納入模式中，期望從兩造談判次數之變化，探究其對於談判結果及談判者之目標達成程度的影響。 本研究以台北港貨櫃儲運中心BOT計畫為實例分析對象，分析結果顯示三種模式皆在第六次談判求得妥協解，模式I之結果為政府每年收取約40.9（百萬元）之分年名目權利金，最優申請人之獲利能力為1.0621，政府財務回收率可達11.689。模式II之結果為政府可收取之權利金約佔每年營收之1.2%，最優申請人獲利能力可達1.0621，政府財務回收率可達11.8324。模式III之結果為政府可向特許公司收取之權利金約為每年運量與0.0000386之乘數，最優申請人獲利能力可達到1.0675，政府財務回收率則可達到11.6567。

The royalty of the BOT projects should be written in BOT concession contract through the concession negotiation according to the Act for Facilitation of Private Participation in Infrastructure Projects (AFPPIP) in Taiwan. In the past, there was few computing formula or negotiation model about royalty for BOT projects to provide public sector or private sectors to negotiate during the biding phase for BOT process. Thus, the public sector cannot judge whether private sector proposed royalty is reasonable or not. The purpose of this study is to develop the royalty negotiation model of BOT projects for government and private sector. This study not only develops royalty negotiation model for BOT projects, but also discusses behaviors of government and private sectors in concession negotiation phase. A bi-level programming model is used to formulate the negotiated royalty problems of BOT projects. The upper level is the government and the lower level is the private sector. Thus, the bi-level programming model is a leader-follower negotiation model. The objection function for the upper level is the maximum of government financial recover ratio (GFRR); and the objection function for the lower level is the maximum of profit index (PI). Also, we establish three models for royalty negotiation model: (1) lump-sum royalty; (2) revenue-based royalty; and (3) ridership-based royalty. In addition, the heuristic algorithm for the bi-level programming model in this study is developed. The heuristic algorithm considers the learning factor, negotiation discount ratio, and negotiation cost about two parties. To illustrate these three models, this study has conducted a case study of Taipei Port Container Logistic BOT Project to find the solution above three models through of the Lingo package and MATLAB programming. The results of this study show that the government and the private sector will get compromise solution about these models at sixth discussion during the contract negotiation. The results are (1) government can charge 40.9 (million) in royalty for lump-sum royalty model, the GFRR is 11.689, and PI is 1.0621. (2) Government can charge 1.2% of revenues in revenue-based royalty model, the GFRR is 11.8324, and PI is 1.0621. (3) Government can charge 0.0000386 of riderships in ridership-based model; and GFRR is 11.6567, PI is 1.0675.