晕啊,对几何分数布朗运动,我在网络上搜索到黎曼积分和wick积分,不同积分的套利结果不一样,然后找到了一篇专门比较这两种积分的文章(见附件),但还是看不怎么明白。有没有大牛形象介绍一下,在黎曼积分下为什么有套利,在wick积分下为什么就没有套利,以及这两种积分联系是什么?
附件最后结论也不明白:
If one is happy with the Wick–Itô–Skorohod definition of a self-financing portfolio then
the fractional Black–Scholes model is free of arbitrage. Let us note, however, that the
proof of the freedom of arbitrage in [3, 8] does not assume that the portfolio is adapted
to the filtration generated by the stock price process. So, in principle one cannot generate
arbitrage even though one knows the future values of the stock. Also, it should be noted
that e.g. the arbitrage opportunity constructed by Rogers [15] does not depend on any
particular notion of integration. The same is true for the pre-limit arbitrage of fractional
Black–Scholes model considered in [19].
On the other hand, under the Riemann–Stieltjes notion of self-financing there is arbitrage
in the fractional Black–Scholes model. So, the questions of fair price of an option or replication
become unclear, even meaningless. Indeed, suppose that there is a minimal hedge
for you favourite European option. Then combining that hedge with an arbitrage strategy
we obtain a super replication with the same initial capital as the (supposed) minimal hedge
has. Thus, it is not reasonable to call the initial wealth of the replicating portfolio the fair
price of the option.
irvingy 发表于 2011-2-19 08:28 
第二,你自己去搜索geometric fractional brownian motion, arbitrage,大把