questions on exotic option pricing algorithms

[Vic_Siqiao]

have u guys read about the book" modeling derivatives in C++" by Justin London? i am now puzzled by the codes on pricing lookback options in his book.

path-dependent options need more parameters in each node, but i just cannt understand the way he use to record the maxima or minima price, through his C++ code, and i eve doubt there may be some mistakes in his examples. i need to test his results by Matlab, although not quite convenient.

so could u guys suggest some other books about pricing algorithms on path-dependent options?

many thanks

 

[Andy Nguyen]

We don't use Justin's book in our program. Take a look at the Fall 2006 required book lis https://www.quantnet.com/threads/required-textbooks-for-fall-2006-courses.27/. One of them is

Implementing Derivative Models (Wiley Series in Financial Engineering)
(Hardcover) by Les Clewlow, Chris Strickland
ISBN: 0471966517

Very good book. Lot of pseudo code meaning you have the frameworks to start, unlike ready-to-run codes from Justin book.

Some of us have that Justin book and just for the record, his codes are buggy and non-compliance to say the least.

 

[Vic_Siqiao]

thanks andy

i think ive already found some mistakes in his codes.i am gonna go correct it.

 

[tikam]

justin london modeling derivatives in C++

Hello Guys,

I agree with Andy, the code is buggy. I have not read the lookback option part,
but for Garch/Interest Rates Models/Monte Carlo simulation, most of the routines fail if you follow the code given by Justin London. Since there is no much explanation, it is very tiedous to correct and even to follow London through is development.

In a way the book is interesting exercice, if you take it like a "correct-the-mistakes" exercice.

I would even say that the book is a con precisely because Justin London claims it works as actual code: for instance, in chapter 10, he gives the figures he used to test one of his routines. So, even with little knowledge of C++, one should be able to implement his code.
I tried, I built it, they ran it: it returns an error (vector out of subscript).

I corrected it (could have been a typing mistake), but then the results the routine returned, they did really not match these claimed by London. Acutally,it did not return any result.