strategy-lab/to_explore/pyquantnews/3_FlowEffects.ipynb

This code downloads historical stock data for TLT, computes log returns, and analyzes mean returns by calendar day of the month. It visualizes these mean returns to identify any calendar effects. The code also simulates a simple trading strategy of buying near month-end and selling at month-start. It evaluates the strategy's performance by aggregating and plotting returns over time. This analysis helps in identifying potential trading opportunities based on calendar effects.

In [ ]:

import pandas as pd
import numpy as np
import yfinance as yf

Download historical stock data for TLT from Yahoo Finance

In [ ]:

tlt = yf.download("TLT", start="2002-01-01", end="2022-06-30")

Compute log returns for the adjusted closing prices

In [ ]:

tlt["log_return"] = np.log(tlt["Adj Close"] / tlt["Adj Close"].shift(1))

Add a column indicating the day of the month for each data entry

In [ ]:

tlt["day_of_month"] = tlt.index.day

Add a column indicating the year for each data entry

In [ ]:

tlt["year"] = tlt.index.year

Group the data by the day of the month and compute the mean log returns

In [ ]:

grouped_by_day = tlt.groupby("day_of_month").log_return.mean()

Plot the mean log returns by calendar day of the month

In [ ]:

grouped_by_day.plot.bar(title="Mean Log Returns by Calendar Day of Month")

Initialize a simple trading strategy of buying before month-end and selling at month-start

In [ ]:

# Set initial returns for the first week of the month to zero
tlt["first_week_returns"] = 0.0

In [ ]:

# Assign log returns to the first week of the month
tlt.loc[tlt.day_of_month <= 7, "first_week_returns"] = tlt[tlt.day_of_month <= 7].log_return

In [ ]:

# Set initial returns for the last week of the month to zero
tlt["last_week_returns"] = 0.0

In [ ]:

# Assign log returns to the last week of the month
tlt.loc[tlt.day_of_month >= 23, "last_week_returns"] = tlt[tlt.day_of_month >= 23].log_return

Compute the difference between last week returns and first week returns to simulate the strategy

In [ ]:

tlt["last_week_less_first_week"] = tlt.last_week_returns - tlt.first_week_returns

Group the data by year and plot the mean returns of the strategy for each year

In [ ]:

(
    tlt.groupby("year")
    .last_week_less_first_week.mean()
    .plot.bar(title="Mean Log Strategy Returns by Year")
)

Group the data by year, compute cumulative sum of returns, and plot it

In [ ]:

(
    tlt.groupby("year")
    .last_week_less_first_week.sum()
    .cumsum()
    .plot(title="Cumulative Sum of Returns By Year")
)

Compute and plot the cumulative sum of returns by day

In [ ]:

tlt.last_week_less_first_week.cumsum().plot(title="Cumulative Sum of Returns By Day")

PyQuant News is where finance practitioners level up with Python for quant finance, algorithmic trading, and market data analysis. Looking to get started? Check out the fastest growing, top-selling course to get started with Python for quant finance. For educational purposes. Not investment advise. Use at your own risk.