1

The duration of a bond is given by : $$ \frac{dP}{dy}\frac{1}{P} = - \frac{D}{1+y} $$ solving for D we have : $$D = -\frac{1}{P}\frac{dP}{dy}.$$ Now we know that the bond price formula is : $$P = \frac{c}{(1+y)^1} + \frac{c}{(1+y)^2}+\cdots + \frac{c+P_p}{(1+y)^T}$$ how can i prove that for zero coupon $c=0$ the duration of a bond is equal the time to maturity for the discrete compounding or it isn't and why? Secondly how the modified duration and convexity adjustment formulas changes when the coupon is zero? Any help would be appreciated since i am new to bond pricing.

1 Answers1

1

When solving for D you forgot to multiply for $1+y$, so that the duration is

$$ D = - \frac{1+y}{P} \frac{dP}{dy} $$

The price of a ZCB is just $P = P_p/(1+y)^T$. The derivative returns $-T P_p/(1+y)^{T+1}$. If you multiply for the other terms almost everything cancels out and you are left with $T$.

To compute the convexity use the price I gave you to compute the second derivative.

Mild consideration: if you have a ZCB, the price is simply given by what you wrote but considering the coupon $0$ ($c = 0$).

Edit [computations]:

The derivative is the derivative wrt the yield (it is the term $dP/dy$). It turns out that:

$$ \frac{dP}{dy} = \frac{-T P_p}{(1+y)^{T+1} }$$ If we multiply this by $-(1+y)/P$ we get:

$$ -\frac{-T P_p}{(1+y)^{T+1} } \frac{1+y}{P} = \frac{T P_p}{(1+y)^{T+1} }\frac{(1+y)(1+y)^T}{P_p} = T$$ as we wanted.

The convexity is instead defined as

$$ C = \frac{1}{P} \frac{d^2P}{dy^2} $$ so we need to compute the second derivative and divide by the price like before

finch
  • 1,611
  • 4
  • 7
  • Your answer is very helpful but can you please provide the calculations i.e the derivative (with to respect to what?) and the multiplication afterwards. And how the MD and the convexity becomes with zcb.Thank you in advance. – Homer Jay Simpson Jul 26 '22 at 13:41
  • I added the computations for the duration and provided the formula for the convexity as requested. hope this clarifies – finch Jul 26 '22 at 13:50
  • So for the convexity i have to do the same procedure? And how the modified duration changes with zcb? Your clarification is splendid.You are more than great and useful.Thanks a lot. – Homer Jay Simpson Jul 26 '22 at 13:56
  • yes, for the convexity use the fomula I gave you: derive once again the derivative of the bond and divide by the price. I don't understand what you mean by "change". the formula are different, just compute them and see what's different about it. – finch Jul 26 '22 at 14:04
  • The second derivative is $Pt\cdot\left(t+1\right)\left(y+1\right)^{-t-2}$ and with simplifications i am resulting $$C = \frac{t(t+1)}{(y+1)^{t+2}}$$.I am new to bond pricing but now i understand why this does not "change" with zero coupon.The right word is i think how it becomes. – Homer Jay Simpson Jul 26 '22 at 14:12
  • Can you confirm? Are my calculations correct? – Homer Jay Simpson Jul 26 '22 at 14:24
  • they seem right to me – finch Jul 26 '22 at 14:35
  • watching again your answer more carefully i noticed that when you say cancel out $P_p$ in the numerator and $P$ in the denominator.Are those quantities the same ? – Homer Jay Simpson Jul 27 '22 at 12:39
  • I wrote it wrong: it is $P_p$ also in the denominator – finch Jul 27 '22 at 13:19
  • ok but the formula of the duration of the bond does not have $P_p$ in the denominator.It has $P$ – Homer Jay Simpson Jul 27 '22 at 13:31
  • I corrected the price of ZCB which was missing an exponent of T. Instead what I said in my previous comment was not true: the computations are all correct. that P is the price and after the first equal sign I inserted the definition of price and performed the computations. don't look what I did: take the formula for D, compte separately all the pieces and then put them together and see if your result matches mine. – finch Jul 27 '22 at 13:41
  • i cannot follow you.I am lost now . – Homer Jay Simpson Jul 27 '22 at 14:03
  • i cannot spot the difference that you say in the exponent.And secondly my question remains : why are you cancelling out two different quantities ? – Homer Jay Simpson Jul 27 '22 at 17:54