def equity(model,
scenario_num=1000,
maxyear=3,
moment_match=False,
frequency=xen.TimeGridFrequency.Day):
if not isinstance(model, xen.Eq1FModel):
raise Exception('Eq1FModel Type is needed')
scenSetID = 'test_set'
scenID = 'equity_test_scen'
resultID = 'equity_test_result'
scenSet = xen.ScenarioSet(scenSetID)
scen = xen.Scenario(scenID, resultID)
scen.general.maxyear = maxyear
scen.general.scenario_num = scenario_num
scen.general.moment_match = moment_match
scen.general.frequency = frequency
# setting additional calculation for martinagle test
model.clear_calc()
calc_randomZ = xen_c.RandomZ()
model.add_calc(calc_randomZ)
calc_uncon_expectation = xen_c.UnconditionalExpectation()
model.add_calc(calc_uncon_expectation)
scen.add_model(model)
scenSet.add_scenario(scen)
scenSet.generate()
res = xen_r.ResultObj(scenSetID, scenID, resultID)
equity = res.get_resultModel(model=model)
equity_uncon_expectation = res.get_resultModel(model=model,
calc=calc_uncon_expectation)
rnd_z = res.get_resultModel(model=model, calc=calc_randomZ)
plt.figure(1)
plt.plot(res.timegrid.data['T'], equity.average(), label='equity_aver')
plt.plot(res.timegrid.data['T'],
equity_uncon_expectation.data[0],
label='equity_analytic')
plt.title('equity_aver vs equity_analytic')
plt.xlabel('time (year)')
plt.ylabel('value')
plt.legend()
plt.figure(2)
plt.title('rand_z')
plt.plot(rnd_z.average())
plt.show()
def rw_shortrate(model,
scenario_num=200,
maxyear=3,
moment_match=False,
frequency=xen.TimeGridFrequency.Day):
if not isinstance(model, xen.Ir1FModel):
raise Exception('Ir1FModel Type is needed')
scenSetID = 'test_set'
scenID = 'test_scen'
resultID = 'test_result_' + model.model_name
scenSet = xen.ScenarioSet(scenSetID)
scen = xen.Scenario(scenID, resultID)
scen.general.maxyear = maxyear
scen.general.scenario_num = scenario_num
scen.general.moment_match = moment_match
scen.general.frequency = frequency
# setting additional calculation for martinagle test
model.clear_calc()
# calc_randomZ = xen_c.RandomZ()
# model.add_calc(calc_randomZ)
calc_uncon_expectation = xen_c.UnconditionalExpectation()
model.add_calc(calc_uncon_expectation)
scen.add_model(model)
scenSet.add_scenario(scen)
scenSet.generate()
res = xen_r.ResultObj(scenSetID, scenID, resultID)
shortrate = res.get_resultModel(model=model)
shortrate_uncon_expectation = res.get_resultModel(
model=model, calc=calc_uncon_expectation)
#rnd_z = res.get_resultModel(model=model, calc=calc_randomZ)
i = 0
plt.figure(1)
plt.plot(res.timegrid.data['T'], shortrate.average(), label='shor_aver')
plt.plot(res.timegrid.data['T'],
shortrate_uncon_expectation.data[0],
label='short_analytic')
plt.title('short_aver vs short_analytic')
plt.xlabel('time (year)')
plt.ylabel('rate (%)')
plt.legend()
plt.figure(2)
plt.plot(res.timegrid.data['T'],
shortrate.average() - shortrate_uncon_expectation.data[0],
label='diff')
plt.title('short_aver - short_analytic')
plt.xlabel('time (year)')
plt.ylabel('rate (%)')
plt.legend()
# plt.figure(3)
# plt.title('rand_z')
# plt.plot(rnd_z.average())
#print('aver Z : ' + rnd_z))
plt.show()
def mc_shortrate(model,
scenario_num=200,
maxyear=3,
moment_match=False,
frequency=xen.TimeGridFrequency.Day):
if not isinstance(model, xen.Ir1FModel):
raise Exception('Ir1FModel Type is needed')
scenSetID = 'test_set'
scenID = 'test_scen'
resultID = 'test_result_' + model.model_name
scenSet = xen.ScenarioSet(scenSetID)
scen = xen.Scenario(scenID, resultID)
scen.general.maxyear = maxyear
scen.general.scenario_num = scenario_num
scen.general.moment_match = moment_match
scen.general.frequency = frequency
# setting additional calculation for martinagle test
model.clear_calc()
# calc_randomZ = xen_c.RandomZ()
# model.add_calc(calc_randomZ)
calc_uncon_expectation = xen_c.UnconditionalExpectation()
model.add_calc(calc_uncon_expectation)
calc_mrk_discount = xen_c.FittingDiscount()
model.add_calc(calc_mrk_discount)
calc_model_discount = xen_c.ModelDiscount()
model.add_calc(calc_model_discount)
scen.add_model(model)
scenSet.add_scenario(scen)
scenSet.generate()
res = xen_r.ResultObj(scenSetID, scenID, resultID)
shortrate = res.get_resultModel(model=model)
shortrate_uncon_expectation = res.get_resultModel(
model=model, calc=calc_uncon_expectation)
mrk_discount = res.get_resultModel(model=model, calc=calc_mrk_discount)
model_discount = res.get_resultModel(model=model, calc=calc_model_discount)
#rnd_z = res.get_resultModel(model=model, calc=calc_randomZ)
print(mrk_discount.data[0])
# result_list['BASE_FITTINGDISCOUNT'].load(1, 1)
#
# shortrate_aver = result_list['BASE_nan'].averave()
# shortrate = result_list['BASE_nan'].load()
# shortrate_uncon_expectation = result_list['BASE_UNCONDITIONALEXPECTATION'].load()
# mrk_discount = result_list['BASE_FITTINGDISCOUNT'].load()
# discount = result_list['BASE_MODEL_DISCOUNTBOND'].averave()
# mrk_spot = result_list['BASE_FITTINGSPOT'].averave()
# mrk_forward = result_list['BASE_FITTINGFORWARD'].averave()
# rnd_z = result_list['BASE_RANDOMZ'].averave()
# print(shortrate_aver)
# print(shortrate[0])
# print(shortrate[1])
# plt.plot(discount)
# plt.plot(mrk_discount)
# plt.plot(mrk_spot)
i = 0
# for z, short_u_exp, short, spot, forward, mrk_disc, disc in zip(rnd_z, shortrate_uncon_expectation[0],
# shortrate_aver, mrk_spot, mrk_forward,
# mrk_discount[0], discount):
# print (
# str(i) + ' : ' + str(z) + ' , ' + str(short_u_exp) + ' , ' + str(short) + ' , ' + str(spot) + ' , ' + str(
# forward) + ' , ' + str(mrk_disc) + ' , ' + str(disc) + ' , ' + str(mrk_disc - disc))
# i += 1
# plt.plot(shortrate_aver)
# plt.plot(shortrate_uncon_expectation[0])
plt.figure(1)
plt.plot(res.timegrid.data['T'], shortrate.average(), label='shor_aver')
plt.plot(res.timegrid.data['T'],
shortrate_uncon_expectation.data[0],
label='short_analytic')
plt.title('short_aver vs short_analytic')
plt.xlabel('time (year)')
plt.ylabel('rate (%)')
plt.legend()
plt.figure(2)
plt.plot(res.timegrid.data['T'],
shortrate.average() - shortrate_uncon_expectation.data[0],
label='diff')
plt.title('short_aver - short_analytic')
plt.xlabel('time (year)')
plt.ylabel('rate (%)')
plt.legend()
# plt.plot(mrk_forward)
# pos = 360
# print(mrk_discount[0][pos])
# print(discount[pos])
#
# print(mrk_discount[0][pos] - discount[pos])
#
# print('disc diff ' + str(pos) + ' : ' + str(mrk_discount[0][pos] - discount[pos]))
plt.figure(3)
plt.title('discount - mrk_discount')
# plt.plot(mrk_discount[0])
# plt.plot(discount)
plt.plot(mrk_discount.data[0] - model_discount.average())
# plt.figure(4)
# plt.title('rand_z')
# plt.plot(rnd_z.average())
#print('aver Z : ' + rnd_z))
plt.show()
def build_test_scen_set():
test_scen_set = xen.ScenarioSet('test_scen_set')
scen_num = 1000
maxyear = 3
moment_match = False
rnd_seed = 10
rnd_skip = 4096 * 2
test_model = xen_s.gbm('test_gbm')
calc_uncon_expectation = xen_c.UnconditionalExpectation()
test_model.add_calc(calc_uncon_expectation)
# for crude
rnd_crude_subtype_list = [
xen.RndSubType.MersenneTwister, xen.RndSubType.Knuth,
xen.RndSubType.Lecuyer, xen.RndSubType.Ranlux3, xen.RndSubType.Ranlux4
]
for rnd_subtype in rnd_crude_subtype_list:
scen = xen.Scenario('scen1', rnd_subtype + '_result')
scen.general.scenario_num = scen_num
scen.general.maxyear = maxyear
scen.general.rnd_type = xen.RndType.Crude
scen.general.rnd_subtype = rnd_subtype
scen.general.rnd_seed = rnd_seed
scen.general.rnd_skip = rnd_skip
scen.general.moment_match = moment_match
scen.add_model(test_model)
test_scen_set.add_scenario(scen)
# for sobol
rnd_sobol_subtype_list = [
xen.RndSubType.Jaeckel, xen.RndSubType.SobolLevitan,
xen.RndSubType.SobolLevitanLemieux, xen.RndSubType.JoeKuoD5,
xen.RndSubType.JoeKuoD6, xen.RndSubType.JoeKuoD7, xen.RndSubType.Kuo,
xen.RndSubType.Kuo2, xen.RndSubType.Kuo3
]
for rnd_subtype in rnd_sobol_subtype_list:
scen = xen.Scenario('scen1', rnd_subtype + '_result')
scen.general.scenario_num = scen_num
scen.general.maxyear = maxyear
scen.general.rnd_type = xen.RndType.Sobol
scen.general.rnd_subtype = rnd_subtype
scen.general.rnd_seed = rnd_seed
scen.general.rnd_skip = rnd_skip
scen.general.moment_match = moment_match
scen.add_model(test_model)
test_scen_set.add_scenario(scen)
# for Halton
scen = xen.Scenario('scen1', xen_r.RndType.Halton + '_result')
scen.general.scenario_num = scen_num
scen.general.maxyear = maxyear
scen.general.rnd_type = xen.RndType.Halton
scen.general.moment_match = moment_match
scen.general.rnd_seed = rnd_seed
scen.general.rnd_skip = rnd_skip
scen.add_model(test_model)
test_scen_set.add_scenario(scen)
# for Faure
scen = xen.Scenario('scen1', xen_r.RndType.Faure + '_result')
scen.general.scenario_num = scen_num
scen.general.maxyear = maxyear
scen.general.rnd_type = xen.RndType.Faure
scen.general.moment_match = moment_match
scen.general.rnd_seed = rnd_seed
scen.general.rnd_skip = rnd_skip
scen.add_model(test_model)
test_scen_set.add_scenario(scen)
return test_scen_set