# Dollar Duration

I am looking over reading 27 in the 3rd book and came across the dollar duration calcs. The book goes into the formula of the following: Say you have a 3 bond portfolio: Column…A…B Bond A…200K…27.50 (mdur) Bond B…500K…10.00 (mdur) Bond C…1,000K…2.50 (mdur) The above weights are market values for argument sake. Lets not worry about convexity at the moment… CFAI formula Sum Products columns A and B and then multiplies by .01 / 3… That makes absolutely no sense since the dollar duration exposure is spread by market value… So by their calcs we get 43,333 as dollar duration, but if you market weight them, the answer is 35,882… Does anyone have any idea on what their problem is? Best regards, Quorky

Doesn’t really help if you don’t provide the example. They divided by 3 because they were par-weighting the portfolio in the book example and not market weighting it. For those who are curious, the example had three \$1mm par bonds in a portfolio with various mkt values and durations.

thanks ymmt.

Sorry, the example is in: Book 3 on p340, example 6… You are correct, they are par weighting the dollar durations. It makes absolutely no sense, because they will be short duration in an interest rate shift… I just dont get it, why try to confuse people? Who would weigh the dollar duration based on par? You would not construct a portfolio with par weights. Why convey this to the candidates? Just got a little upset because it makes me wonder, what have I read that I dont know well and have taken in without questioning it… hmmmm