針對不動產抵押貸款風險而言，此貸款會有違約與提前清償的風險，因此，貸款收益率會反映貸款的風險。除此之外，不動產價值與借款人和金融機構關係也會影響貸款的條件，並且再進一步影響不動產抵押貸款價值。在不動產抵押貸款的相關實證研究文獻中，都發現借款人與金融機構關係會影響不動產抵押貸款價值與借款成本；當金融機構在承作不動產抵押貸款時，貸款人與金融機構之間的互動關係是會影響借貸條件，由於過去許多建構不動產抵押貸款模型研究，他們並未將貸款人與金融機構關係納入模型中來考慮，所以本研究主要是把借款人與金融機構關係建構在不動抵押貸款模型。本研究利用隨機過程模型來結合借款人與金融機構之互動關係，並且以數值分析法來模擬借款人與金融機構之關係，對不動產抵押貸款借款收益率的影響。根據數值分析發現，當借款人與金融機構關係越好，則金融機構會給予借款人較低的借貸成本；若借款人與金融機構關係較普通，則在承作不動產抵押貸款時，其借款人的借貸成本可能比較高。本研究所建立的模型數值結果與金融機構關係實證研究的相關文獻一致，所以可以提供金融機構，使用本研究模型來研究不動產抵押貸款的價值與貸款收益率分析。

We study the potential model instability problem with respect to mortgage
default risk and prepayment risk, therefore the loan yield will reflect the risk. In addition, the real estate value and the financial institution relationship will also affect the loan condition,and will further influence mortgage value. Past research model do not make the financial institution relationship in the mortgage model, therefore this research is construction the financial institution relationship in the mortgage model. We find that about the mortgage empirical study literature, discovered that the financial institution relationship can affect the mortgage value and the loan yield. When financial institution makes the mortgage, the financial institution relationship will affect the lending condition, as a result of the past many construction mortgage model study, they do not integrated the financial institution relationship with the mortgage model. Our study use stochastic process model and integrated with the financial institution relationship. We use numerical methods to simulate the mortgage value with financial institution relationship. According to numerical analysis, when the financial institution relationship are better, then the financial organ possibly produces lowly mortgage cost; If the financial institution
relationship are not well, when receives makes the mortgage, it loan cost to be possibly higher. The result of our study is consistent with the other empirical study's literature, therefore we can use this model to research the mortgage and the loan yield analysis.