以往已經有不少以實質選擇權評估房地開發專案的研究,但這些研究仍有許多假設並不洽當,因此本研究的實質選擇權採用以下假設:(1) 土地成本採用具有隨機性的機會成本。(2) 折現率採用必要報酬率。(3) 房地產之總銷售金額視為有趨勢之隨機變數。(4) 不設定履約期限。此外,配合其他不同假設,本文提出三種評價模式:(1) 土地成本與房地產總銷售金額分成兩個期望值比例無關的隨機變數 (2) 土地成本與房地產總銷售金額分成兩個期望值比例固定的隨機變數 (3) 土地成本與房地產總銷售金額合為單一隨機變數。並以二叉樹法計算這些模型。結果顯示:(1) 房地產總銷售金額現值越高、房地產價格波動率越高、房地產價格變化率越高、履約價格(建築成本)越低、存續期間越長、土地價格變化率與房地產價格變化率相關係數越低、土地現在價值越低、必要報酬率越低,則擴張NPV越大。(2) 無論哪一個模式,最重要的變數都是房地產總銷售金額、房地產價格變化率、履約價格(建築成本)、土地現在價值;存續期間、相關係數、土地價格波動率的影響都很小。(3) 三種模式中,專案的擴張NPV以模式一最高,模式二略低,模式三則低很多。
In the past there have been numerous studies that employed the real options approach to evaluate the value of real estate development projects, but there are some inappropriate assumptions in these studies. Hence, the real options approach in this study employs the following assumptions: (1) the cost of land employ the opportunity cost with randomness. (2) The discount rate employs the required rate of return. (3) The total sales amount of real estate is a random variable with trend. (4) There is no predetermined expired time on the real estate development projects. In addition, using various other assumptions, this study proposed three evaluation modes: (1) the cost of land and real estate total sales amount are two random variables without fixed ratio between them. (2) The cost of land and real estate total sales amount are two random variables with fixed ratio between them. (3) The cost of land and real estate total sales amount is considered as a single random variable. And corresponding binomial tree methods are proposed to solve these models. The results showed that: (1) the higher the real estate total sales amount, the higher the volatility of real estate prices, the higher the rate of change in real estate prices, the lower the strike price (construction cost), the longer the period of existence of the right, the lower the correlation coefficient between the change rate of land cost and the change rate of real estate prices, the lower the current cost of the land, the lower the required rate of return, then the greater the extended NPV. (2) No matter which model, the most important variables are the real estate total sales amount, the volatility of real estate price, the strike price (construction cost), and the current cost of the land; the least important variables are the period of existence of the right, the correlation coefficient, and the volatility of the land cost. (3) Regarding to the extended NPV, model 1 is the highest. Model 2 is slightly lower, and Model 3 is much lower than Model 1.