本研究針對三種不同架構的境外結構性產品進行評價和風險分析，第一檔為連結股權之境外結構型商品，Fixed Coupon Note (FCN)，為亞洲金融市場中交易量最大的結構型商品架構；第二檔為連結指數之境外結構型商品，Bearish Sharkfin，為看空型的保本商品架構；第三檔為連結匯率之境外結構型商品，Dual Currency Note (DCN)。 首先分析各架構目標投資者和特色，再以Black-Scholes模型在幾何布朗運動的假設下，透過Itô’s Lemma計算其中隨機微分方程，推導出單資產及多資產隨機變動過程模型，並考慮到產品條款、波動率、無風險利率、匯率和相關係數…等因素，採用蒙地卡羅方法對相關資產的隨機過程進行建模，計算所有模擬結果的預期收益，同時考慮市場風險和發行機構之信用風險，以無風險利率加上發行機構之信用違約交換(CDS)作為折現率，將所有模擬收益折現至期初，從而確定目標產品的理論價值。 最後利用模擬結果，統計出每種產品中不同情境之發生機率及預期收益，將所有收益結果由小至大排列，用於計算風險值(VaR) 和條件風險值(CVaR)，估算特定機率下的最大損失，以及損失超過特定VaR條件時的平均損失。;This study concentrates on the valuation and risk analysis of three distinct offshore structured product types: Fixed Coupon Notes (FCN) linked to equities, which is the most popular type of structured product in Asia; Bearish Sharkfin, a principal-protected structured product linked to an index; Dual Currency Notes (DCN) related to foreign exchange. Initially, the study analyzes the target investors and characteristics of each product, assuming the underlying price processes follow Geometric Brownian Motion and considers various factors such as product terms, volatility, risk-free rates, exchange rates, and correlation coefficients. The Monte Carlo simulation method is employed to model the stochastic process of the underlying assets, the expected returns from all sim-ulation outcomes are calculated, factoring in both market and issuer risks. The risk-free interest rate, augmented by the Credit Default Swap (CDS) rate of issuers, is applied as the discount rate to bring all expected returns to initial price, determining the theoretical value of the products. Utilizing the simulation results, the study calculates the probabil-ity of occurrence and expected returns for different scenarios within each product. Ul-timately, all potential outcomes are organized from the least to the most significant and are used to compute the Value at Risk (VaR) and Conditional Value at Risk (CVaR), these measures are designed to estimate the maximum loss at a given probability and the average loss when the loss exceeds a specific VaR condition.
