Design, Analysis, and Implementation of Efficient Financial Risk Management Algorithm

瀏覽統計 Email 通知 RSS Feed

簡歷

基本資料

Project title

Code/計畫編號

NSC102-2221-E001-015-MY3

Translated Name/計畫中文名

財務風險管理演算法的設計、分析、與實作

Funding Organization/主管機關

National Science and Technology Council

Co-Investigator(s)/共同執行人

何建明(計畫主持人)
吳牧恩

Department/Unit

Institute of Information Science (IIS)

Website

https://www.grb.gov.tw/search/planDetail?id=11264899

Year

2015

Start date/計畫起

01-08-2015

Expected Completion/計畫迄

31-07-2016

Co-Investigator(s)

William Wei-Yuan Hsu

Bugetid/研究經費

836千元

ResearchField/研究領域

資訊工程--硬體工程

在過去五年裡，發生了多次重大的經濟災難，包括了2007-2008年的次級房貸風暴、2010-2011年的歐債危機、以及2012年的美國財政懸崖。在這些金融風暴中，銀行和金融機構蒙受了巨大損失。這些事件突顯出金融體系中風險管理的重要性，其關鍵是在效率與彈性中取得平衡。對此，國際金融監管體系已經修訂和發展日益嚴格的法規，包括Basel I-III，IFRS，Solvency II 等，藉以提高金融機構的透明度，並確保其健全運作。然而，相關文獻則指出這些金融風暴的成因包含許多政治和商業因素。其中包含過度倚賴信評機構的制度性問題，因此，最好的解決方案，則有賴投資機構或個人能夠提昇自己的了解和掌控風險的能力。本計畫將聚焦在以下幾項風險管理技術相關議題上。首先，過去的信評模型未能即時實際反應金融機構和公司的信用評等。第二，投資組合的價值在金融海嘯時期嚴重被錯估，代表現有的定價模式仍有改善空間。第三，考慮更多的投資資產及變數，伴隨了大量的計算需求。在這個三年期計畫中，第一年我們將研究一些財務上的風險管理議題，包含基金評價、資產定價、投資理論、以及均衡結算等。我們也將研究設計或改良相關演算法，以提升風險管理的計算效能。在第二年，我們著重在套利交易的研究，包含了評價模型的建立以及找尋新的配對交易方法。我們也將探討動量交易策略，包括找出市場趨勢，並預測交易的時機和風險。此外，我們也將研究如何評估交易演算法與共同基金的價值。最後，在本計畫的第三年，我們將實現過去兩年的研究成果，建構實用的金融評價與交易風險估算平台。我們也將進一步探討這項平台的可能應用，包括協助使用者了解和掌控投資風險等。我們也將探討運用多核運算或雲端計算等技術來提昇計算的效率。 In the past five years, there were several large-scale economic catastrophes in the world such as Subprime mortgage crisis in 2007-2008, Euro-zone crises in 2010-2011, and US fiscal cliff in 2012. During these crises, banks and financial institutions around the world have suffered huge losses. It highlights the importance of risk management in balancing efficiency and resilience in the financial system. The international financial supervision systems have revised and developed increasingly rigorous regulations, including Basel I-III, International Financial Reporting Standard (IFRS), Solvency II, etc., to improve soundness and transparency of operations of financial institutions. There are political and business reasons leading to this crisis including the over-reliance on the rating results of the rating institutions. The best strategy, as recommended by the economists, is to improve awareness and ability for the investors, institutions or individuals, to better understand and control their own financial risk. We will thus focus on the following technical issues on understanding and control of the risk of investment strategies. First, rating technology mush be improved to timely and accurately capture company defaults. Second, misprice of portfolios during the crisis period shows that existing pricing models are problematic. Third, the demand of computing power increases as more factors and more assets are taken into account to achieve better risk diversification. In this three years project, in the first year, we will start with designing and improving algorithms in portfolio management including fund evaluation, asset pricing, investment theory, and equilibrium accounting. In the second year, we will focus on studying risk management of trading strategies including arbitrage trading in which a pricing model of asset or security is developed and further tested for mean reversal property. We’ll also study momentum trading strategy in which algorithms are developed to prospect future trend of a market and to predict the timing and risk of trading. In the third year, we’ll realize our research results in the past two years to construct a computing framework of credit rating and evaluation of trading risks. We’ll also explore potential applications of the framework including identifying users who are interested in understanding and controlling his/her investment risk. To cope with the high complexity involved in rating and risk measurement, we will also design efficient algorithms based on multi-ore and cloud computing models to improve computing efficiency.

DSpace CRIS

Design, Analysis, and Implementation of Efficient Financial Risk Management Algorithm

基本資料

Description