2
$\begingroup$

For a vanilla bond, as coupon goes down ,absolute duration goes up, but absolute dv01(absolute change in price for a 1bp increase in rates ) goes down. so ... what is the message to take away from this... which is riskier!

Edit:

Perhaps the answer to my question is , that it depends on what matters to you: your % return on investment, or your dollar P&L. For % ROI , duration is the measure to use , and for dollar P&L, DV01 is the one to use.

I would think that traders would care more about dollar P&L since the traders job is to hedge the market risk over the life of the trade , so my follow up q is - why do Bond traders look at duration rather than DV01 ?

$\endgroup$
2
  • 1
    $\begingroup$ I think absolute DV01 should be read as the USD increase in price for a 1bps lowering of the market yield. With smaller flows, absolute price and absolute price-change gets smaller. $\endgroup$
    – Mats Lind
    Commented Feb 20, 2017 at 15:05
  • $\begingroup$ The risk is measured not by DV01 but by $\frac{DV01}{P_0}$ $\endgroup$
    – Alex C
    Commented Feb 20, 2017 at 17:54

2 Answers 2

1
$\begingroup$
  1. Per USD invested, the lower coupon bond is risker.
  2. Per USD face value, the higher coupon bond is riskier.

2.) is trivial because the higher coupon bond constitutes of the original bond plus a series of positive coupons. The two has the same sign of their its risk, i.e. they lose present value with higher yield. So the higher coupon bond is riskier, it has the risk of the lower coupon bond plus the risk of the extra coupons.

Looking at 1.) and partitioning it into the repayment and the coupons; consider that that the repayment is riskier per amount invested as it is longer than the coupons. Then a reallocation towards the repayment through decreasing the coupons is a reallocation towards the riskier constituent, hence lower coupons increse the relative risk.

$\endgroup$
3
  • $\begingroup$ can you elaborate / explain further , eg what do you mean by "... sign of its risk" ? $\endgroup$
    – Randor
    Commented Feb 21, 2017 at 11:57
  • $\begingroup$ Yes, please let me edit my post to make this clearer! /Mats $\endgroup$
    – Mats Lind
    Commented Feb 22, 2017 at 9:02
  • $\begingroup$ aah, i understand your explanation now, thanks :) $\endgroup$
    – Randor
    Commented Feb 26, 2017 at 9:21
1
$\begingroup$

Edit: I've got a chance to talk with a risk manager. The conclusion is that DV01 is used more by traders simply because they care more about the actual dollar value. Portfolio managers may use duration instead because they get compensated in a different way.

Some relevant discussions are available on this Wikipedia page.


According to Wikipedia, dollar duration and DV01 are essentially the same thing. Dollar duration is a common risk measure.

Your question "why do Bond traders look at duration rather than DV01 ?" is a little mis-postulated. Bond traders do look at DV01 aka dollar duration. Other durations (Modified, Macauley, etc.) are different concepts.

$\endgroup$
4
  • $\begingroup$ yes i am sure bond traders these days also look at dv01. $\endgroup$
    – Randor
    Commented Feb 21, 2017 at 18:32
  • $\begingroup$ unfortunately with phone i cannot edit prev comment! i meant to say that the q is, why at all would a trader look at duration , and key rate duration too of course . i am guessing the answer is a) historical reasons - as duration , comvexity, redington immunisation, were the original ways of measuring and hedging bond risk $\endgroup$
    – Randor
    Commented Feb 21, 2017 at 18:36
  • $\begingroup$ and also, that these 2 risk measures really are not so substantially different i think... $\endgroup$
    – Randor
    Commented Feb 21, 2017 at 18:37
  • $\begingroup$ @Randor i've updated my answer $\endgroup$
    – Will Gu
    Commented Feb 23, 2017 at 15:22

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.