Estimating the error of a linear regression

Excel offers the possibility to calculate the coefficients a and b of a linear regression line f(x): y=a+bx with one click. Where can I find a "HowTo" for calculating in Excel the estimated error for a certain value f(x) including the contribution from the covariance ?

This is one of those questions where I think a good understanding of the underlying statistical theory is needed before worrying about how to implement in Excel. Unfortunately, I don't have the expertise to give you that information. With that in mind, the LINEST function (assuming you are using LINEST) can output supplemental statistics that include the standard error in each coefficient and the standard error in y. I expect if we understood the statistics behind what you want, some combination of these values could be likely used to determine the overall error in your regression.

It looks like not being possible with Excel to calculate the "correct" estimated error for an estimated result calculated with the regression line. It would require the correlation matrix which Excel does not seem to calculate. Is there a VBA program which could be used together with Excel to perform such an analysis ?