QuotePlus sometimes, the practical solutions that do work are so ugly mathematically that you prefer not to expose them Everyone knows the equations; it's the undocumented features (e.g. "we do this and that and the other when the parameter becomes bla bla in order to avoid bbb", ) that people want to see A good example was the Cheyette model discussion.

QuoteOriginally posted by: jawabeanQuoteOriginally posted by: mjwriting good code would make their book more popular, no doubt. It would also delay the launch for about another 3 years...dont they have their stuff implemented already?! dont they run some kind of code to test what they write about?Cuchulainn explained it perfectly. if it's not the code, then it could be pseudo code. for instance, Knuth's books come with machine code, it cant be run anywhere(*), but it's so useful.Clelow and Strickland is a good example of good pseudocode, it is very helpful. But I'm afraid that the issues discussed in the book will be too specific for pseudocode to be of great value.

First I really appreciate the authors' contributions on this book. As discussed before, if this book will be with code or not, it depends on the author's focus. If this is just a pure theoretical book on mathematical proof, it is fine. Otherwise this one turns out to be a practical handbook, code part is very very important. As the author mentioned, most of codes have been developed in some IB environment and all codes owned by IB. Probably it is hard to show up all of them in the book due to some copyright issues. As Cuchulainn said, a lot of quants are working on programming during their working time. That is really true. I guess only a very small part of them focus on pure theoretical proof or new initials development. When I came across a model, I will always ask myself what it is the first step if I want to implement, what inputs are, what algo is in order to make it work in PC. For super quants, probably the theoretical part is enough. But for the other quants on the way like me, I think code part is really really important as well since from the code part, we can understand the model itself deeply. My dreaming book seem to be with easy words, reproduced numerical results, steps on how to generate results, explanation on model's input. Anyway, thank you for your guys' great work!!!Max

I would like to see insights and inferences - on model dynamics, real-life issues (techniques), extrapolation problems etc. I expect that is what this book will do. There are already several books on the models and C++ gurus.

just some remark re usefulness of codes:1. very useful to see how people actually code. as you can pick up ideas on how to do certain things better.if you only code by yourslef yes you learn a lot, but you might be doing things in an overcomplicated way all your life without ever noticing it.Seeing it done better would help. How can you know that what you re doing is overcomplicated otherwise?2. Also, it s all nice to write a paper and show nice results. But without codes how do you know how much is hidden?For instance there is this paper by Vladimir on calibration of the CEV model for FX, neither me or my colleagues have been able to reprdocude the results in the paper. And there are many posts by other people on forums encoutering the same problems and wasting time trying to find where they went wrong.Please do not see (2) as an attack - I have found papers by Vladimir very helpful. I m just arguing that codes can have value.Now of course, I can also see that I wouldn t want to give away my codes:1. They would use value, I would use value.2. And when you explain something to somebody, you want the person to understand, not to just "repeat" without having undestood. 3. And of course "repeating" without understanding (using codes that are not undersood), could be potentially dangerous.

PS: great news about the book, looking fwd to reading it!

Thank you!Btw added some blurb on all the chapters here here

Dear Vladimir,just saw that the book comes in three volumes. Anything about the pricing yet...?Looking fwd to the book(s).Best wishes,Lapsi

Havent thought much about pricing yet. what would people find a reasonable price to be?Vladimir

Hi Vladimir,that is a nice problem. Something like pricing an exotic product. One might take the expectation of all books on IR modelling and add a margin ;-)On the other hand side there might be liquidity problems (e.g. students, academics, ...)I really do not know. As cheap as possible but as expensive as necessary. I know this does not help at all!Since I know that such a big project costs work measured in months / years the price is too cheap anyway!Best Lapsi

I would think that the upper boundary will be at the risk books - bankers won’t mind at this rate but it will limit all important audience - the University students (who can be tempted around 80).

Ok so this Risk book has 350 pages and is by the authors I don't think I recognize. it costs 135 pounds.Or this 250 page book for £125. again, not so familiar with the authorwell I think we will be rather cheaper than that (per page)V

QuoteOriginally posted by: piterbargHavent thought much about pricing yet. what would people find a reasonable price to be?Vladimir$30 - and I'll consider buying it. otherwise, i'll take it from library and ... memorize it

Well, I would compare it with Mercurio's book (GBP 85) and Hughstron's old book (GBP 145) in terms of pricing (in the risk publication). Of course, other classical texts (Brigo, Li, Rebonato etc) will be compared against this which are priced well below risk books. More realistically speaking, I will compare these with classics in terms of content.

Quote$30 - and I'll consider buying it. otherwise, i'll take it from library and ... memorize itCheapskate.