完全自償BOT案財務可行性分析中違約風險評估模型之建構－－以水力電廠為例

Value-at-risk (Var) Analysis of Profitability Indices for BOT Projects– With Case Study of Water Power Plants

國內重大的民間參與國家公共工程建設案，如台灣高鐵案、高雄捷運案、台北101 大樓，於興建期因資金無法到位、工程逾期完成，或於營運期營收不如預期等種種問題，使得BOT 計畫浮現出虧損的可能性。對國內BOT 案最大融資者之國內聯貸銀行而言，聯貸銀行將有可能受到因對BOT 案投資不當，致使聯貸銀行承受鉅額損失之風險。BOT 案是屬於專案融資，專案融資案計畫複雜度高、計畫期長，不同於一般傳統借貸，銀行所承受風險之可能性及複雜度均變高，進而影響銀行對計劃的融資意願。因此高鐵案、高捷案、中正機場捷運案於期初財務分析均顯示可行，但事後於執行階段卻發現問題重重，顯示當時風險分析不夠精確深入。所以可見完整量化的風險分析及管理策略是專案融資案重要研究課題。本研究建立一套專案財務分析模型，實際分析水力電廠案例，考慮最佳出資比後，做為最佳化財務結構的標準組。其次再利用標準組數據，針對11 項風險因子，進行敏感度分析，求出第一階段會影響計劃財務可行性指標之關鍵財務影響因子並加以排序，結果發現建設成本、電力年產量及電價Uav(得標時的報價)這三種為最重要財務管理因子。再進行第二階段之三種情境模擬與第三階段加入機率條件進行蒙地卡羅模擬計算專案的風險大小。本研究案例經蒙地卡羅模擬分析結果，六大評估指標之累計分佈機率函數圖(CDF)在於其限制條件下，違約機率最大值約在13.75%左右，同時本模式可以計算出各獲利指標之風險值(VaR)。本研究建立一套適用專案融資之違約機率評估模型(DRM)，模式可以分析BOT 案財務因子對六大財務可行性指標的影響，排列出財務因子影響順序及影響程度，並決定出關鍵財務影響因子，可以做為後續BOT專案執行時風險管理之重點，期能降低BOT 案風險。另外模式也可建構評估指標量化機制、分析專案風險大小、評估專案的違約機率、獲利指標風險值。這些將專案風險量化的資訊，可使得BOT 專案計畫更形穩健，讓BOT 案三大參與者政府、投資者與銀行團之間達到三贏。

There are many mega-scale BOT projects in Taiwan, such as Taiwan High Speed Rail project and the Kaohsiung MRT projects. These projects encountered some problems such as the time delay or cost overrun in construction phase, which can lead to higher project failure. This outcome may induce a large amount of loss to the loan providers. Project financing is normally adopted by BOT projects. Mega-projects are characterized by long concession period and engineering complexity in project. Hence, the bankers provide the loans with high risk imbedded. A completed risk analysis is essential to the participants of the BOT projects. We adopt the financial model in the literature to serve as a basic model for financial simulation. We conduct a sensitivity analysis for determining the critical factors in the model which can become the focus for management in the implementation of the projects. Afterwards, an in-depth interview is conducted with the experts to determine the optimum, pessimist, and most likely value for the financial factors as well as the possibility for the optimum and the pessimist value of the financial factors. Based on relevant information, a scenario analysis and Monte Carlo simulation can be performed. With the aid of @risk, the default risk of the project can be calculated and the results show that the maximum default risk is around 13.75%. A default risk model (DRM) for calculating PI-VaR is constructed in this paper which can provide a powerful tool to estimate the impact of the financial factors on the profitability indices of the projects. By comparing the degree of the impact on each profitability index, the critical factors of the model can be determined accordingly. Relevant information of risk quantification is very useful for the managerial consideration and, this model can provide the quantified information for the investors, public agents, and banks in their decision making for the BOT projects.