<t>Is there a specific matrix formula defining the composition of a portfolio B (several assets) that can be use to hedge a portfolio A ?ThanksPS: I know the case in the case of one asset versus another one for instance with a stock versus an index future but my question in more general, is there a ...

Do you mean you use some kind of local volatility in each of your tree's nodes ?

<t>Don't worry Mr Quantman, it was a kind of joke because my thread didn't seem to catch attention except for marvellous Rouletabille II..I don't make mystery of what I use now.To answer to your question I was pretty disappointed by direct regressions, (pb of colinearity, stability, ...), so you can...

Hi Quantman,I did not see your approach in this thread ..?

Your answer here:http://www.math.nyu.edu/research/carrp/ ... ingoh1.pdf

I think that a price impact function can be adquate, but it requires more than just a tick by tick database, I mean at least a five limite orders book...

In a nutshell, everytime you go long for a call, convertible bond, forward, etc.. on astock you "sell" the dividend..

Did you search toward Stochastic dividends Scheme?

YES I found what I had searched for...a little more complicated than previous expectations but great results...I will be a rich man...

http://www.wspc.com.sg/profiles/anncat/ ... 000085.pdf

Collaterally, can we see the difference between a stock price (random walk) and one of its Exponential Moving Average as a Ornstein Ulhenbeck process ?If so, what can you say of the difference of two EMA ?

<t>Quotef) You talk about a "very short term horizon (max 1 day)" I wonder if you are doing some cash-and-carry-like arbitrage ...It could be the case, but you can think of a wayto hedge some special exotic basket option , etc...I take special attention to your point e)and according to a pure statis...

<r>You can see at :<URL url="http://www.nag.co.ukhttp//www.olin.wustl.edu/faculty/zhou/clz5.pdfhttp://www.anderson.ucla.edu/acad_unit/finance/wp/2001/7-01.pdf"><LINK_TEXT text="http://www.nag.co.ukhttp://www.olin.wus ... 1/7-01.pdf">http://www.nag.co.ukhttp://www.olin.wustl.edu/faculty/zhou/clz5.pdf...

Define your Half-life It might be easier to tune the decayfactor if it's specified as a half-life, i.e the time required for an old value's contribution to shrink to half. might be something like: -log(.5)/half

<t>Exactly Rouletabille,But you can also imagine that you can't trade on set A because of their lack of liquidity , or internal constraint from your Risk Management (black or red list), or because stocks of set A are halted (suspended)... The horizon is very short term (max. one day) In case of a re...