Bootstrap Term Structure Model

Overview

A bootstrapped term structure is a term structure where the yields are various points along the curve are recursively measured, and then a curve is fit to those points. The method of fit is typically a cubic spline. (see curve estimation)

Curve Types

The bootstrap.py file can bootstrap the following curves, given a set of datapoints. (inlcuding with a set of spreads)

  • PiecewiseLogLinearDiscount
  • PiecewiseLogCubicDiscount
  • PiecewiseSplineCubicDiscount
  • PiecewiseLinearZero
  • PiecewiseFlatForward
  • PiecewiseLinearForward

Sample Code

data = {"date": "2026-08-10", "1M": 3.79, "1.5M": 3.8, "2M": 3.84, "3M": 3.89, "4M": 3.91, "6M": 4.0, "1Y": 4.04, "2Y": 4.25, "3Y": 4.31, "5Y": 4.41, "7Y": 4.56, "10Y": 4.72, "20Y": 5.25, "30Y": 5.25} data2 = [] for key in data: if key != 'date': item = [key, data[key]] data2.append(item) pass data3 =ty.term_data(data2) curve = ts.BootstrappedCurve(type=ts.BootstrapTypes.PiecewiseLogLinearDiscount,day_count=ty.DAY_COUNT, data=data3)

Contents