2007年是中華電信企業民營化的第二年，如同以往地在業務面及財務面締造了佳績。然而，在2007年9月時，中華電信與香港高盛簽下一紙長達10年期的外匯避險合約，使得中華電信在隔年第一季公佈高達新台幣40億元的未實現匯兌損失，引起市場一片譁然與投資人的輿論，認為此合約讓收入多以新台幣計價的中華電信陷入了不必要的衍生性商品陷阱之中。本個案讓閱讀者以當時中華電信財務長謝劍平的角度，帶領閱讀者解構、分析這個極具爭議的的結構型商品。過程中牽涉到避險決策、契約評價、代理問題、公司治理、外匯預測、市場效率性等議題。2008年10月契約觸及匯價32.7終止，中華電信最終獲利新台幣3010萬元出場，新台幣10多億的帳面損失也全數回沖。個案閱讀者能從中學習到重要的思考過程，並做出權衡之下對股東利益最好的決策，我們發現這個外匯避險契約本身並無絕對好壞，因為這是一個風險控管與節省成本的抵換關係，端看決策者的出發點、需求及風險偏好決定。 / 2007 was the second year after the privatization of Chunghwa Telecom, the financial performance was strong as usual. However, on September of the same year, Chunghwa Telecom had signed a 10-year currency hedging contract with Goldman Sachs, which leads to NTD 4 billion unrealized book losses in the first season of 2008. Investor and the public were shocked about it, because it was not compatible with the company’s corporate image, and the book losses was too huge. Many public opinion criticized that it was a big mistake for Chunghwa Telecom, which let them fall into the trap of financial derivatives. This case put students back to the scenario of September 2007, played the role as Chunghwa Telecom’s CFO, CP Shieh. The case will guide them understand how to decompose and analyze a tailor made structure product provided by investment bank. In the analyzing process, readers will involve many interesting issues, like hedging decision, contract valuation, agency problem, corporate governance, forex forecasting and market efficiency. October 2008, the contract knocked out by reaching NTDUSD quote 32.7. Chunghwa Telecom end up gained NTD 30.1 million from the contract and over NTD 1 billion unrealized book losses were also reversed. Case readers can acquire important intuition during the analyzing process and make decision based on the shareholder’s best interest. There is no absolute answer to dictate whether this hedging contract was good or not, because it was a tradeoff between risk control and cost saving. It depends on the decision maker’s point of view, demand and risk preference.

本篇論文共分為兩部分，第一部份是以每年交易量非常大的外匯選擇權（FX Option）市場以及台指選擇權為例，以Brigo 及Mercurio這兩位學者於2000年提出的Lognormal Mixture model (簡稱LM model)為基礎，捕捉選擇權市場中典型的波動度微笑（Volatility smile）曲線之特性。第二部份係商品評價之應用，是以大陸地區發行的匯率連動結構型商品（Structure Notes）為主。 第一部份中我們分別採用LM 模型（Lognormal Mixture Model）、Shifting LM模型（Shifting Lognormal Mixture Model）及LMDM模型（Lognormal Mixture with Different Mean Model）等三種模型，用以衡量其實際上在外匯選擇權市場及台指選擇權中波動微笑曲線校準的準確性。結果顯示LM模型、Shifting LM模型及LMDM模型均能有效地反應並捕捉出選擇權市場中波動度微笑曲線之特性，而其中又以LMDM模型的效果最佳，其無論在波動度校準或是選擇權價格評價上的誤差均最小。 第二部分是以「中國銀行匯聚寶0709G掛鉤美元兌加元匯率之加元產品」的匯率連動結構型商品為例，以Garman and Kohlhagen（1983）外匯選擇權模型求出其封閉解並作發行商期初利潤分析，然後再用蒙地卡羅模擬法進行投資人期末報酬分析。此外，亦針對此種商品的敏感性與避險參數作分析。

Submitted by Cristiane Shirayama (cristiane.shirayama@fgv.br) on 2011-06-03T16:41:12Z No. of bitstreams: 1 66080100251.pdf: 1071566 bytes, checksum: 633248672cb6ac94f704bfeda06b29d3 (MD5) / Approved for entry into archive by Vera Lúcia Mourão(vera.mourao@fgv.br) on 2011-06-03T16:46:36Z (GMT) No. of bitstreams: 1 66080100251.pdf: 1071566 bytes, checksum: 633248672cb6ac94f704bfeda06b29d3 (MD5) / Approved for entry into archive by Vera Lúcia Mourão(vera.mourao@fgv.br) on 2011-06-03T17:00:17Z (GMT) No. of bitstreams: 1 66080100251.pdf: 1071566 bytes, checksum: 633248672cb6ac94f704bfeda06b29d3 (MD5) / Made available in DSpace on 2011-06-03T18:49:55Z (GMT). No. of bitstreams: 1 66080100251.pdf: 1071566 bytes, checksum: 633248672cb6ac94f704bfeda06b29d3 (MD5) Previous issue date: 2011-02-11 / Nos últimos anos, o mercado brasileiro de opções apresentou um forte crescimento, principalmente com o aparecimento da figura dos High Frequency Traders (HFT) em busca de oportunidades de arbitragem, de modo que a escolha adequada do modelo de estimação de volatilidade pode tornar-se um diferencial competitivo entre esses participantes. Este trabalho apresenta as vantagens da adoção do modelo de volatilidade estocástica de Heston (1993) na construção de superfície de volatilidade para o mercado brasileiro de opções de dólar, bem como a facilidade e o ganho computacional da utilização da técnica da Transformada Rápida de Fourier na resolução das equações diferenciais do modelo. Além disso, a partir da calibração dos parâmetros do modelo com os dados de mercado, consegue-se trazer a propriedade de não-arbitragem para a superfície de volatilidade. Os resultados, portanto, são positivos e motivam estudos futuros sobre o tema. / In recent years, the Brazilian option market has grown considerable, especially with the emergence of the High Frequency Traders (HFT) in search of arbitrage opportunities, so that the appropriate choice of a volatility estimation model should become a competitive differentiator among these participants. This paper presents the advantages of adopting the Heston stochastic volatility model on the construction of the volatility surface for the Brazilian US Dollar option market, as well as the easiness and the computational gain by applying the Fast Fourier Transform technique on the models differential equations resolution. Furthermore, from calibration of the model parameters to market data, it is possible to bring the no-arbitrage property to the volatility surface. The results, therefore, are positive and motivate further studies on the subject.