The question below is from topics test Fixed Income Portfolio Management - Monts. I know this has been discussed in the previous years but still not clear on this. Based on the answer, sounds like they only look at the spread duration instead of duration. So,
My questions are:
What is the relationship between spread duration and tracking error?
What is the relationship between duration and tracking error?
I got this same one wrong as well and was scratching my head at first but now I understand. To answer your questions:
Think about everything in relation to the benchmark. If you have a higher spread duration contribution to industrials versus the benchmark then that means if spreads move in industrials then your tracking error will increase (in simplistic terms your return will deviate from the benchmark because of this mismatch).
Similar to the above. Think about if you run a portfolio with a duration of 10 versus a benchmark with a duration of 1 (although I know this is probably not an appropriate benchmark). If rates go up in a parallel fashion then your portfolio will likely do much worse (i.e. high tracking error). Inverse is true if rates go down in a parallel fashion. In either case the duration mismatch is high and so is the tracking error.