Duplicate Curve Pillars in Swap Bootstrapping
Summary
The document explains a QuantLib curve-building error that occurs when multiple instruments map to the same maturity pillar. In that situation, the bootstrap has more constraints than a single zero rate can satisfy, particularly if the instruments imply conflicting values.
It illustrates the issue with a forward rate agreement and a swap whose maturities overlap at a curve tenor. The proposed remedy is to change the instrument set—for example, use a later-maturity swap—so the bootstrap can infer the intervening zero rates from the available market instruments. The discussion is a brief response to one reported CAD swap valuation error, not a general walkthrough of QuantLib setup or a proof that this adjustment is appropriate for every curve.
Key ideas
- Multiple instruments assigned to the same curve pillar can overdetermine a bootstrap rate.
- Overlapping instrument maturities may produce conflicting implied values for one zero rate.
- Changing the swap maturity can remove a duplicate pillar and allow intervening rates to be bootstrapped.
- The suggested adjustment is an example and should be checked against the instruments and curve design in use.
Tags
Full text
# Quantlib : How to resolve ' more than one instrument with pillar' in valuing swaps? # Quantlib : How to resolve ' more than one instrument with pillar' in valuing swaps? I had valued interest rate swaps of most of the currencies keeping my valuation date as 13th Sep 2019. But I faced a problem of 'RuntimeError: 2nd leg: more than one instrument with pillar December 18th, 2019'while valuing CAD.3M. Any suggestions to resolve this type of error or where to look in the Quantlib library. ## Answer by user35980 (score 1) https://quant.stackexchange.com/a/54852 This error occurs because you have multiple instrument maturities overlapping for a single zero tenor, so the bootstrapping procedure has an overdetermined number of solutions (likely with conflicting values) for a given zero rate. For example, if you use a 12x15 FRA and a 1y swap rate in curve construction. The correction in this example will be to use a 2y swap instead (so the curve will bootstrap the intervening zero rates implied from the 12x15 FRA and the 2y swap).
Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.