0
$\begingroup$

how common is it to use the US STRIP market directly to bootstrap a par yield curve? I am having quite a few problems using cubic (peicewise) splines on coupon bearing yield curve, but I see a fair few prices on bbg for STRIPs so wanted to know if this is an equivalent/acceptable method.

i do understand obviously it is not as liquid as coupon bearing market and that itself can have some implications, but what is the magnitude of difference in liquidity? are we talking 100m vs 1tn or 50bn vs 75bn type numbers?

$\endgroup$
10
  • $\begingroup$ What are the "a few problems" specifically? Generally speaking STRIPS aren't used for building coupon curves. $\endgroup$
    – Helin
    Commented Jul 29, 2023 at 0:16
  • $\begingroup$ @Helin thanks for coming back. Let me describe in a bit of detail: $\endgroup$
    – user67825
    Commented Jul 29, 2023 at 8:39
  • $\begingroup$ @Helin I have a list of every treasury issue that is in existence (coupon bearing). to keep things simple i am using issues post 2022 to calibrate my yield curve. I am using excel solver to fit the curve with different peicewise cubic polynomials (similar to tuckmans method described) at knot points. I have adjusted for specials to keep them in the curve (the adjustment is small anyway). The fit is bad (in excess of 10bps of errors between model / market for instruments in the calibration). Further, bonds not in the calibration set have errors > 100bps. $\endgroup$
    – user67825
    Commented Jul 29, 2023 at 8:49
  • $\begingroup$ just to be clear - i have also tried this with even more recent issues, i.e. just OTRs (this remediates the problem of the calibrated instruments pricing errors) - but older bonds have some errors in excess of 100bps still (therefore, I thought I should include more bonds in the calibration). $\endgroup$
    – user67825
    Commented Jul 29, 2023 at 11:34
  • $\begingroup$ That sounds like an implementation issue, not an input issue. This shouldn't happen. I recommend that you check your methodology/code. $\endgroup$
    – Helin
    Commented Jul 29, 2023 at 16:38

0

Your Answer

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

Browse other questions tagged or ask your own question.