QuantLib 金融计算——一个使用 ActualActual
时需要注意的陷阱
ActualActual
是分析债券时最常用的 day counter(天数计算规则),根据 StackExchange 上的讨论(https://quant.stackexchange.com/questions/12707/pricing-a-fixedratebond-in-quantlib-yield-vs-termstructure),在使用 ActualActual
时最好附加上债券现金流支付的日期表(Schedule
对象),否则在计算贴现因子的时候可能产生偏差。
然而,这种操作隐藏了一个陷阱,下面用一个案例来解释。
import QuantLib as ql
print(ql.__version__)
today = ql.Date(10, ql.November, 2020)
ql.Settings.instance().evaluationDate = today
effectiveDate = ql.Date(21, ql.May, 2019)
terminationDate = ql.Date(21, ql.May, 2029)
tenor = ql.Period(1, ql.Years)
calendar = ql.China(ql.China.IB)
convention = ql.Unadjusted
terminationDateConvention = convention
rule = ql.DateGeneration.Backward
endOfMonth = False
schedule = ql.Schedule(
effectiveDate,
terminationDate,
tenor,
calendar,
convention,
terminationDateConvention,
rule,
endOfMonth)
scheduleEx = ql.Schedule(
effectiveDate,
ql.Date(21, ql.May, 2031),
tenor,
calendar,
convention,
terminationDateConvention,
rule,
endOfMonth)
dayCounter = ql.ActualActual(
ql.ActualActual.Bond, schedule)
print('results 1:')
print(dayCounter.yearFraction(today, today + ql.Period(8, ql.Years)))
print(dayCounter.yearFraction(today, today + ql.Period(9, ql.Years)))
print(dayCounter.yearFraction(today, today + ql.Period(10, ql.Years)))
dayCounterEx = ql.ActualActual(
ql.ActualActual.Bond, scheduleEx)
print('results 2:')
print(dayCounterEx.yearFraction(today, today + ql.Period(8, ql.Years)))
print(dayCounterEx.yearFraction(today, today + ql.Period(9, ql.Years)))
print(dayCounterEx.yearFraction(today, today + ql.Period(10, ql.Years)))
'''
1.20
results 1:
8.0
8.526027397260274
8.526027397260274
results 2:
8.0
9.0
10.0
'''
如果计算涉及到的日期超过了日期表的范围,即 Schedule
前两个参数确定的范围,日期计算可能产生与预期不一致的错误。
一个简便有效的解决方法是为债券对象和 ActualActual
分别提供各自的 Schedule
对象,ActualActual
的 Schedule
对象范围更大一点,以便包括所有可能涉及的日期,就像案例中的做法一样。