本研究は, 金融機関や一般企業にとって不可欠な, 金融リスクのより良いマネジメントを行うために必要な数値計算技術の開発を行うことが主たる目的であった。これまでの金融実務では, 金融リスクの定量的分析を行うときには, リスク資産価格のリターンの正規性を想定していることが多かったが, 本研究では実証分析の結果明らかとなった非正規性が金融リスク管理にどのような影響を与えるかに関する分析を行った。また, その非正規性の想定の下, 主に準モンテカルロ法による数値計算技術を用いて正確かつ効率的な定量か手法を提案し, その実用性を明らかにした。
The goal of this research was to develop a new series of computational methods that can help management handle financial risk more efficiently. Most existing numerical methods used in current financial institutions are based on a classical assumption that returns of financial assets, interest rates, and exchange rates are under normal distribution. However, many empirical studies indicate that historical financial data do not support the assumption and we need more accurate models. Motivated by these findings, we first investigated the effects of new models on practical risk management in terms of valuation, hedging, and risk measure computation. Then, we have proposed a couple of simulation methods that can generate samples from any class of distribution. We have also proposed an enhanced quasi-Monte Carlo method that can evaluate financial instruments as well as financial risk measures more accurately and more efficiently.
|
