<t>HiTo caclulate estimated correlation I use univariate garch model to estimate volatilities with for each serie of return:sigma_n=gamma+alpha* u_n-1^2+beta*sigma_n_1^2.cov_n=gamma+alpha*x_n-1 y_n-1+beta*cov_n-1Then i calculate corr(X,Y)_n=cov_n/(sigmaX_n*sigmaY_n)1°) I found that in the Hull. Do y...

<t>Quote1) The best way is to get someone to make a market on a "vanilla" European basket option and back out implied correlation.2) Think of correclation as being related to volatility (variance). How would you model volatility in the absence of tradeable data? That is, if you only have historical ...

<t>does anybody have an idea about how you can get an implied correlation between two stocks?1°) if they are included in the same index 2°) and especially if they are not.actually, i m working on strucured notes pricing and i would like to get as input an implied correlation matrix instead of histor...

ok. no problem.but it seems you need a basket option with citi and toyota to get an implied correlation. But I heard it xas possible without it. because it s hard to find it for all equitities.

<t>QuoteOriginally posted by: daveangelif you have the vol for exchange rate X1 and X2 (both against the same base) and the vol the cross (X12) then relationship is given by the cosine rule:vol12^2 = vol1^2 + vol2^2 + 2*rho*vol1*vol2hence rho = (vol12^2 - vol1^2 - vol2^2)/(2*vol1*vol2)let me summari...

<t>QuoteOriginally posted by: daveangel1. Index variance versus constituent stock variance aka dispersion trading2. implied vol from cross fx options versus the pairsthanks daveangelIt would be very nice if you could explain the procedure to extract implied correlation, especially for 2°) having imp...

<t>From What kind of product it would be possible to extract implied corraletion between two products. and what would be the procedure?I heard it was possible but i don't find anything on it. (The only thing i found is how to calculate the implied correlation (a scalar) of an index but it doesn't se...

<t>QuoteOriginally posted by: AaronThe sample covariance is not an unbiased estimate of the true covariance, nor are the sample variance or standard deviation unbiased estimates of the population parameters.For a simple example, suppose we have three equally likely possible outcomes, X = 0, Y = 0; X...

<t>thx and I agree with you but it doesn't answer my question.It was:the correlation coefficient estimator calculated from the unbiaised covariance estimator divided by the product of unbiaised standard error estimator is an unbiased estimator of correlation coefficient or not??and if yes, how do we...

ok so if I have IBM, L'oreal, and Mitsubishi in my basket, do I have to put everything in EUR. And just making my pricing with the equities correlation matrix (so without crossing currencies). and the problem of currencies correlation is now ok???

QuoteOriginally posted by: daveangelit all depends what your payoff is ...For instance a worstOfCould you tell it would be considered and if you use EUR/USD USD/EUR About what point one have to be careful?

<t>mmmh...sounds not a bad idea..but your suggestion is for each date t to consider prices in AUR for instance with the exchange rate of date t (with EUR) and thento calculate returns and then calculate the matrix correlation used after for pricing.But often prices are taken in their money=> returns...

<t>hello,If I don't make any mistake, empirical covariance matrix is an unbiaised estimator of the real covariance matrix. It is the same for (unbiaised) empirical std deviation for each random variable.My question is: Do we have the same property for the correlation matrix i.e. Is the associated co...

<t>Hello,I have to price a structured product, so i need to consider the correlation matrix of my stocks.But I have international stocks in my basket. Do you know how I have to introduce that in my correlation matrix. for instance, if I have usd jpn and european stocks, is it better to use EUR/USD o...

hello,I heard that it could be useful to estimate an Igarch model on asset returns in order to correct data for the correlation matrix calculation.does anybody have an idea about the utility of such a method??thx and have a good day