QuantLib: BondFunctions.zSpread no coincide limpio precio "exactamente"

Question

Estoy usando QuantLib para calcular los precios de los bonos de tasa fija en diferentes escenarios. En el primer paso me gustaría replicar el precio actual de mercado mediante el ajuste de la curva de rendimiento de la zspread. Utilizando la función de BondFunctions.zSpread puedo acercarme (en el ejemplo de abajo (97.85 frente a un precio de mercado de 98), pero me pregunto por qué nosotros no convergen al mismo precio cuando la alteración de la propagación de forma continua. Me estoy perdiendo algo?

El código siguiente se adapta Simple QuantLib de Bonos de Matemáticas

import pytest
import QuantLib as ql

from QuantLib import *

# Construct yield curve
calc_date = Date(1, 1, 2017)
Settings.instance().evaluationDate = calc_date

spot_dates = [Date(1,1,2017), Date(1,1,2018), Date(1,1,2027)]
# corrected!
# spot_rates = [0.0, 0.04, 0.04]
spot_rates = [0.04, 0.04, 0.04]

day_count = SimpleDayCounter()
calendar = NullCalendar()
interpolation = Linear()
compounding = Compounded
# corrected!
compounding_frequency = Annual
compounding_frequency = Semiannual
spot_curve = ZeroCurve(spot_dates, spot_rates, day_count, calendar,
                       interpolation, compounding,
                       compounding_frequency)

spot_curve_handle = YieldTermStructureHandle(spot_curve)

# Construct bond schedule
issue_date = Date(1, 1, 2017)
maturity_date = Date(1, 1, 2022)
tenor = Period(Semiannual)
calendar = NullCalendar()
business_convention = Unadjusted
date_generation = DateGeneration.Backward
month_end = False

schedule = Schedule(issue_date, maturity_date, tenor, calendar,
                    business_convention, business_convention,
                    date_generation, month_end)

# Create FixedRateBond Object

coupon_rate = 0.05
coupons = [coupon_rate]
settlement_days = 0
face_value = 100

fixed_rate_bond = FixedRateBond(settlement_days,
                            face_value,
                            schedule,
                            coupons,
                            day_count)

# Set Valuation engine
bond_engine = DiscountingBondEngine(spot_curve_handle)
fixed_rate_bond.setPricingEngine(bond_engine)

# Calculate present value
value = fixed_rate_bond.NPV()
assert value == pytest.approx(104.49, abs=1.e-2)


# fix a hypothetical market price
px = 98.

# compute the implied z spread
zspread = ql.BondFunctions.zSpread(fixed_rate_bond,
                                   px,
                                   spot_curve, day_count, compounding,
                                   compounding_frequency, calc_date, 1.e-16, 1000000, 0.)


def impl_clean_price(spread):
    spread1 = ql.SimpleQuote(spread)
    spread_handle1 = ql.QuoteHandle(spread1)
    ts_spreaded1 = ql.ZeroSpreadedTermStructure(spot_curve_handle,
                                                spread_handle1)
    ts_spreaded_handle1 = ql.YieldTermStructureHandle(ts_spreaded1)
    ycsin = ts_spreaded_handle1
    fixed_rate_bond = FixedRateBond(settlement_days,
                                    face_value,
                                    schedule,
                                    coupons,
                                    day_count)
    # Set Valuation engine
    bond_engine = DiscountingBondEngine(ycsin)
    fixed_rate_bond.setPricingEngine(bond_engine)
    value = fixed_rate_bond.cleanPrice()
    return value

# the two clean prices are 98 and 97.8517891975
print px
print impl_clean_price(zspread)
print abs(px-impl_clean_price(zspread))

Answer 1

En caso de no proporcionar la capitalización de la convención de a ZeroSpreadedTermStructure de forma explícita, se tendrá en cuenta el pasado repartidas de forma continua componiendo y aplicarlo a la base de la curva en consecuencia.

Usted tendrá que crear una instancia de la curva de lugar como:

    ts_spreaded1 = ql.ZeroSpreadedTermStructure(spot_curve_handle,
                                                spread_handle1,
                                                compounding,
                                                compounding_frequency)

Esto le dirá a la ZeroSpreadedTermStructure cómo interpretar la propagación de la cita y el resultado será el correcto precio de los bonos.