119 lines
3.2 KiB
Python
119 lines
3.2 KiB
Python
|
import datetime
|
||
|
|
import math
|
||
|
|
import random
|
||
|
|
|
||
|
|
# import time
|
||
|
|
from typing import List
|
||
|
|
|
||
|
|
from dateutil.relativedelta import relativedelta
|
||
|
|
|
||
|
|
import pyfacts as pft
|
||
|
|
|
||
|
|
|
||
|
|
def create_prices(s0: float, mu: float, sigma: float, num_prices: int) -> list:
|
||
|
|
"""Generates a price following a geometric brownian motion process based on the input of the arguments.
|
||
|
|
|
||
|
|
Since this function is used only to generate data for tests, the seed is fixed as 1234.
|
||
|
|
Many of the tests rely on exact values generated using this seed.
|
||
|
|
If the seed is changed, those tests will fail.
|
||
|
|
|
||
|
|
Parameters:
|
||
|
|
------------
|
||
|
|
s0: float
|
||
|
|
Asset inital price.
|
||
|
|
|
||
|
|
mu: float
|
||
|
|
Interest rate expressed annual terms.
|
||
|
|
|
||
|
|
sigma: float
|
||
|
|
Volatility expressed annual terms.
|
||
|
|
|
||
|
|
num_prices: int
|
||
|
|
number of prices to generate
|
||
|
|
|
||
|
|
Returns:
|
||
|
|
--------
|
||
|
|
Returns a list of values generated using GBM algorithm
|
||
|
|
"""
|
||
|
|
|
||
|
|
random.seed(1234) # WARNING! Changing the seed will cause most tests to fail
|
||
|
|
all_values = []
|
||
|
|
for _ in range(num_prices):
|
||
|
|
s0 *= math.exp(
|
||
|
|
(mu - 0.5 * sigma**2) * (1.0 / 365.0) + sigma * math.sqrt(1.0 / 365.0) * random.gauss(mu=0, sigma=1)
|
||
|
|
)
|
||
|
|
all_values.append(round(s0, 2))
|
||
|
|
|
||
|
|
return all_values
|
||
|
|
|
||
|
|
|
||
|
|
def sample_data_generator(
|
||
|
|
frequency: pft.Frequency,
|
||
|
|
num: int = 1000,
|
||
|
|
skip_weekends: bool = False,
|
||
|
|
mu: float = 0.1,
|
||
|
|
sigma: float = 0.05,
|
||
|
|
eomonth: bool = False,
|
||
|
|
) -> List[tuple]:
|
||
|
|
"""Creates TimeSeries data
|
||
|
|
|
||
|
|
Parameters:
|
||
|
|
-----------
|
||
|
|
frequency: Frequency
|
||
|
|
The frequency of the time series data to be generated.
|
||
|
|
|
||
|
|
num: int
|
||
|
|
Number of date: value pairs to be generated.
|
||
|
|
|
||
|
|
skip_weekends: bool
|
||
|
|
Whether weekends (saturday, sunday) should be skipped.
|
||
|
|
Gets used only if the frequency is daily.
|
||
|
|
|
||
|
|
mu: float
|
||
|
|
Mean return for the values.
|
||
|
|
|
||
|
|
sigma: float
|
||
|
|
standard deviation of the values.
|
||
|
|
|
||
|
|
Returns:
|
||
|
|
--------
|
||
|
|
Returns a TimeSeries object
|
||
|
|
"""
|
||
|
|
|
||
|
|
start_date = datetime.datetime(2017, 1, 1)
|
||
|
|
timedelta_dict = {
|
||
|
|
frequency.freq_type: int(
|
||
|
|
frequency.value * num * (7 / 5 if frequency == pft.AllFrequencies.D and skip_weekends else 1)
|
||
|
|
)
|
||
|
|
}
|
||
|
|
end_date = start_date + relativedelta(**timedelta_dict)
|
||
|
|
dates = pft.create_date_series(start_date, end_date, frequency.symbol, skip_weekends=skip_weekends, eomonth=eomonth)
|
||
|
|
values = create_prices(1000, mu, sigma, num)
|
||
|
|
ts = list(zip(dates, values))
|
||
|
|
return ts
|
||
|
|
|
||
|
|
|
||
|
|
market_data = sample_data_generator(num=3600, frequency=pft.AllFrequencies.D, skip_weekends=False)
|
||
|
|
mts = pft.TimeSeries(market_data, "D")
|
||
|
|
print(mts)
|
||
|
|
|
||
|
|
# print("Datediff=", (mts.end_date - mts.start_date).days)
|
||
|
|
# stock_data = sample_data_generator(num=3600, frequency=pft.AllFrequencies.D, skip_weekends=False, mu=0.12, sigma=0.15)
|
||
|
|
# sts = pft.TimeSeries(stock_data, "D")
|
||
|
|
# print(sts)
|
||
|
|
|
||
|
|
# start = time.time()
|
||
|
|
# alpha = pft.jensens_alpha(
|
||
|
|
# asset_data=sts, market_data=mts, risk_free_rate=0.052, return_period_unit="months", return_period_value=1
|
||
|
|
# )
|
||
|
|
|
||
|
|
# print(alpha)
|
||
|
|
|
||
|
|
# print("Alpha calculation took", time.time() - start, "seconds")
|
||
|
|
|
||
|
|
# print("Correlation=", pft.correlation(sts, mts))
|
||
|
|
|
||
|
|
rr = mts.calculate_rolling_returns(frequency="D")
|
||
|
|
|
||
|
|
print(117, rr[rr.values < 0.1])
|