以下是引用lgc_2001在2006-5-28 22:41:00的发言: 回:
Feynman-Kac 是用概率论给出PDE的解,是研究PDE的诸如Dirichlet边值问题的概率基础。
拜托!其不是数值计算,跟数值计算无关!在金融上其是鞅方法解决Black-Scholes理论的基础!
建议你随便找本随机分析看看吧!
一再强调,Monte Carlo 跟PDE理论无关,但可以解决PDE的某些数值计算,其主要作用是随机模拟!
应用Monte Carlo 可以直接模拟出underlying asset的演化路径,从而不给出方程,也能近似得到期权价值
建议你去看看徐钟济编的“模特卡罗方法”上海科学技术出版社,1985
Noboday said Feynman-Kac is a numerical method. It just relates SDE and PDE and Monte Carlo is a numerical method to solve SDE.
For the BS PDE, you can back out the risk neutral measure using Feynmac-Kac and then use Monte Carlo. I think this IS a numerical method to solve the PDE.
建议你随便找本随机分析看看吧!
- I don't like this, but as you did it, I sugget you read some books in stochastic calculus and PDE, specifically, Shreve, Bjork, Tavella and Randall, and Lipton. I'm not a math guy and I'm only interested in books that are related to finance. Maybe you can find some pure math books that support you, but I seriously doubt it.
建议你去看看徐钟济编的“模特卡罗方法”上海科学技术出版社,1985
- What's so special about this book? Does it say that you cannot solve the BS PDE by Monte Carlo? For Monte Carlo, I suggest Jackel and Glasserman, although they don't spend much on the thereom. Again, I seriously doubt that the book you mentioned is superior to my suggestions.
[此贴子已经被作者于2006-5-29 6:10:58编辑过]