003.snowball.py

import datetime as dt
import pandas as pd
import numpy as np
from dateutil.relativedelta import relativedelta
from dal import *

spot = 1.00
vol = 0.15
rate = 0.0
div = 0.0
ko = 1.00
ki = 0.88
coupon = 0.069
today = dt.date(2023, 3, 1)
maturity = dt.date(2025, 3, 1)

event_dates = []
event_dates = ["KI", "KO", "STRIKE", "COUPON"] + [Date_(today.year, today.month, today.day)]
start = event_dates[4]

EvaluationDate_Set(start)
events = [f"{ki:.6f}", f"{ko:.6f}", f"{spot:.6f}", f"{coupon:.6f}", "alive = 1 is_ki = 0"]

n_paths = 2 ** 20
use_bb = False
rsg = "sobol"
freq = "1M"

event_dates.append(f"START: 2023-06-01 END: 2025-02-01 FREQ: {freq}")
events.append(
    f"""
    if spot() < KI:0.001 then is_ki = 1 end
    if spot() > KO:0.001 then call pays alive * COUPON * DCF(ACT365F, {start}, PeriodEnd) alive = 0 end
    """)

event_dates.append(Date_(maturity.year, maturity.month, maturity.day))
events.append(
    f"""
    if spot() < KI:0.001 then is_ki = 1 end
    if spot() > KO:0.001 then call pays alive * COUPON * DCF(ACT365F, {start}, {event_dates[-1]}) alive = 0 end
    call pays alive * is_ki * (spot() - STRIKE) + alive * (1.000000 - is_ki) * COUPON * DCF(ACT365F, {start}, {event_dates[-1]})
    """
)

print("------   Model Parameters  ------")
print(f"rate : {rate* 100:.2f}%")
print(f"div  : {div * 100:.2f}%")
print(f"vol  : {vol * 100:.2f}%\n")

print("------ Product Description ------")
print(f"NPV date  : {event_dates[4]}")
print(f"Maturity  : {event_dates[-1]}")
print(f"knock in  : {ki:.2f}")
print(f"knock out : {ko:.2f}")
print(f"# of obs  : {(len(event_dates) - 4) * 12 if freq == '1M' else (len(event_dates) - 4) * 51}")
print(f"# of paths: {n_paths}\n")


product = Product_New(event_dates, events)
model = BSModelData_New(spot, vol, rate, div)

print("------ Product Evaluation  ------")
now = dt.datetime.now()
res = MonteCarlo_Value(product, model, n_paths, rsg, use_bb, False)
all_res = {"Non-AAD": [res["PV"], np.nan, np.nan, np.nan, np.nan, (dt.datetime.now() - now).total_seconds() * 1000]}

now = dt.datetime.now()
res = MonteCarlo_Value(product, model, n_paths, rsg, use_bb, True)
all_res["AAD"] = [res["PV"], res["d_spot"], res["d_vol"], res["d_rate"], res["d_div"], (dt.datetime.now() - now).total_seconds() * 1000]

df = pd.DataFrame.from_dict(all_res)
df.index = ["NPV", "delta", "vega", "dP/dR", "dP/dDiv", "Elapsed (ms)"]
print(df.T.to_markdown())