Timeline for Why does the ultra long-end of a yield curve invert?
Current License: CC BY-SA 4.0
6 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Aug 11, 2019 at 13:22 | comment | added | dm63 | Well to simplify, consider only zero coupon bonds with a yield y and maturity T: $ Price, P=e^{-yT}, dv01 = dP/dy = -Te^{-yT} and convexity = d^2P/dy^2 = T^2 e^{-yT}$. Hence the convexity per unit dv01 is proportional to T. | |
| Aug 11, 2019 at 8:59 | comment | added | quanty | Thanks @dm63 Would you mind posting thats maths that make it a mathematical fact? | |
| Aug 10, 2019 at 2:59 | comment | added | dm63 | @attack68 If you do this trade with 10s30s what you find is that 10s tend to move more than 30s, which offsets the effect. Whereas the market tends to move 30s and 40s in parallel. | |
| Aug 9, 2019 at 14:14 | comment | added | Attack68♦ | This argument should work with 10s30s as well as 30s50s, no? In general the potential value of the roll-down (carry) and (under-priced) convexity should also be taken under advisement in relation to recent (or opportunistic) price levels and overall value at risk (delta) of the trade. | |
| Aug 9, 2019 at 10:09 | comment | added | Chris Taylor | Other answers kind of hint at it, but this one is the clearest - if the curve wan't inverted at the ultra-long end then you could make put on carry, positive convexity trades and nearly always make a profit. | |
| Aug 9, 2019 at 0:17 | history | answered | dm63 | CC BY-SA 4.0 |