運用實質選擇權分析軟體開發專案的風險

Abstract

傳統軟體工程把軟體專案視為是一種工程活動，以符合預算內準時交付優質的軟體作為軟體專案管理的目標。所以傳統軟體工程觀點的風險管理考慮的是如何避免或降低不確定性對品質，時程與預算的影響。對傳統軟體工程而言，不確定性是不好的，要設法避免或降低。對現代企業而言，軟體專案不只是一種工程活動，也是一種商業活動。企業從事商業活動的目的是要獲利。從財務理論的觀點來看，不確定性未必是不好的，只要有適當的風險溢酬(risk premium)，不確定性的結果可以從造成損失的風險轉變為創造獲利的機會。所以軟體專案的風險管理除了可以採用軟體工程觀點，以避免風險造成負面影響作為管理的策略外；也可以採用財務理論的觀點，以計算積極承擔風險應當獲取的風險報酬作為管理的策略。 本論文旨在研究如何將軟體工程與財務理論的方法整合，運用風險溢酬的觀念將軟體估算的不確定性反映在軟體專案的價格與與風險管理的決策上，以風險溢酬的觀念管理軟體專案的風險。研究的問題包括： - 以實質選擇權的觀念分析風險控制策略。 - 以選擇權定價模型量化軟體專案的風險。 傳統使用預期貨幣值分析法估算風險曝露值來評估風險控制策略效益的方式不能反映軟體成本波動的風險。這也是為什麼PMBoK的風險管理知識領域建議使用蒙地卡羅法估算風險曝露值。但是蒙地卡羅法在使用上有計算耗時且複雜的問題，在實用時會有因使用成本導致效率不佳的問題。因此，本論文期望能夠證明在軟體開發專案的風險管理上，採用選擇權的觀念分析風險控制策略，並且使用Black-Scholes選擇權定價模型估算風險曝露值是一個更為簡單與有效率的方法。

A software project is regarded as an engineering activity from a traditional software engineering perspective. The primary goal of the software project management is to deliver quality software on time within budget. Therefore, the objective of the risk management is to prevent or reduce the impact to quality, schedule or budget caused by uncertainties. The uncertainties thus should be prevented or reduced from a traditional software engineering perspective. From business point of view, a software project is not just an engineering activity, but also a business activity. The goal of a business activity is to make profits. Uncertainties are not necessarily bad from a financial engineering perspective. This is because given an adequate risk premium, we could turn the risks caused by uncertainties to an opportunity to profit. Therefore, in addition to reducing the project uncertainties using appropriate tools for project management, we could also aggressively undertake the uncertainties and turn them into profits using financial engineering theories. The thesis studies how to integrate the software engineering concept with the financial engineering theories to measure the uncertainties of the software project estimation using the risk premium concept. The objectives of the studies include: - Analyze the risk control strategy using the real option concept. - Measure the risk of software project using the Option Pricing Model. The Expected Monetary Value Analysis is widely used in the cost-benefit analysis of the risk control strategy. However, the analysis cannot measure the volatility of software cost estimation, and this is why Monde Carlo Simulation is recommended by PMBoK to measure the risk exposure number in Project Risk Management. But the Monde Carlo Simulation is sometimes over-complicated and usually requires large computing resource. Therefore, in this thesis we apply the real option concept to analyze the risk control strategy, and use Black-Scholes Option Pricing Model to measure the risk exposure number, hoping to prove that this is a simpler and more efficient way to cope with risk management in software project development.