Monte Carlo LSM (Least Squares Method)

[egondragon]

Hi all

I have some questions about Monte Carlo pricing using Longstaff & Schwartz method :
* What is the purpose of the LSM exactly?
* What is the curse of dimension means ? do you know?
* Once Monte Carlo used to price my options, which methods can i use to optimize a portfolio, for example, if i want to build a delta neutral portfolio ?

I have read that Monte Carlo method is related to dynamic programming, i do not really cope to that, for me this is two different domains.

Here is an example of these concepts :
I am a investor
My wealth is w0
the risk free rate is r
my expected return is R = exp (rT)
The time horizon is T

The risky stock with a price of S0

My strategy is a buy and hold one
no borrowing
no short selling

The processing is this :
1. Generate k scenarios with a proba = p(k), this part use Monte Carlo
2. multiply each value by sigmaT
3. This give us a norm sample
4. Compute the excessive return
5. Compute the excess utility
6. Maximize this utility

No dynamic programming in it I thought
I do not search for a precise response, but if anybody hears about this, it will be helpful

Thanks