Comparing US stonks ($VTI) and World ex-US ($ACWX) for a couple of months before and after the start of 2025 like the FT article. Code assumes Yahoo! Finance CSV files are available.

analyze.py

import pandas as pd
import numpy as np
import pylab as plt
from pandas.tseries.offsets import MonthEnd

plt.style.use('ggplot')
plt.ion()

# Load the datasets
us_data = pd.read_csv('vti.txt.csv', parse_dates=['Date'], index_col='Date')
world_ex_us_data = pd.read_csv('acwx.txt.csv',
                               parse_dates=['Date'],
                               index_col='Date')

# Ensure both datasets cover the same date range
start_date = max(us_data.index.min(), world_ex_us_data.index.min())
end_date = min(us_data.index.max(), world_ex_us_data.index.max())
us_data = us_data.query("@start_date <= index <= @end_date")
world_ex_us_data = world_ex_us_data.query("@start_date <= index <= @end_date")

# Calculate daily returns
us_data['Daily Return'] = us_data['Close'].pct_change()
world_ex_us_data['Daily Return'] = world_ex_us_data['Close'].pct_change()

# Combine the daily returns into a single DataFrame
returns = pd.DataFrame({
    'US': us_data['Daily Return'],
    'World_ex_US': world_ex_us_data['Daily Return']
}).dropna()


# Resample to monthly periods and calculate Pearson correlation for each month
def monthly_pearson(df):
    return df.corr().iloc[0, 1]


monthly_corr = returns.groupby([returns.index.year,
                                returns.index.month]).apply(monthly_pearson)

# Convert the grouped index to a datetime index
monthly_corr.index = [
    pd.Timestamp(f'{year}-{month:02d}') + MonthEnd(0)
    for year, month in monthly_corr.index
]

# Plot the monthly Pearson correlation coefficients
plt.figure(figsize=(14, 8))
plt.plot(
    monthly_corr.index,
    monthly_corr.values,
    label='Monthly Pearson Correlation',
)
plt.title(
    'Monthly Pearson Correlation Between US (VTI) and World ex-US (ACWX) Stock Indices'
)
plt.xlabel('Year')
plt.ylabel('Pearson Correlation Coefficient')
plt.grid(True)
plt.tight_layout()
plt.savefig('pearson.png', dpi=300)
plt.savefig('pearson.svg')

print(monthly_corr.describe())
print(monthly_corr.sort_values()[:10])

fix = lambda ser: ser / ser[-1] * 100
year_months = [(10, 2024), (11, 2024), (12, 2024), (1, 2025), (2, 2025),
               (3, 2025)]

plt.figure(figsize=(14, 8))
for m, y in year_months:
    idx = np.logical_and(us_data.index.year == y, us_data.index.month == m)
    plt.plot(
        us_data.index[idx],
        fix(us_data['Close'][idx]),
        label='US',
        c='r',
    )
    plt.plot(
        world_ex_us_data.index[idx],
        fix(world_ex_us_data['Close'][idx]),
        label='World ex-US',
        c='b',
        linestyle='-.',
    )

handles, labels = plt.gca().get_legend_handles_labels()
plt.legend(handles[:2], labels[:2])
plt.grid(True)
plt.xlabel('Date')
plt.ylabel('Index')
plt.suptitle(
    'US (\$VTI, red solid) vs World ex-US (\$ACWX, blue dashed), monthly rebased'
)
plt.tight_layout()
plt.savefig('monthly.png', dpi=300)
plt.savefig('monthly.svg')

fig, axs = plt.subplots(len(year_months), 1, sharex=True, figsize=(14, 8))
for ax, (m, y) in zip(axs, year_months):
    idx = us_data.index > pd.Timestamp(f"{y}-{m:02}-01")
    ax.plot(
        us_data.index[idx],
        fix(us_data['Close'][idx]),
        label='US',
        c='r',
    )
    ax.plot(world_ex_us_data.index[idx],
            fix(world_ex_us_data['Close'][idx]),
            label='World ex-US',
            c='b',
            linestyle='-.')
    ax.grid(True)
    ax.set_ylabel('Index')

axs[-1].legend()
axs[-1].set_xlabel('Date')

fig.suptitle(
    'US (\$VTI, red solid) and World ex-US (\$ACWX, blue dashed), rebased')
fig.tight_layout()
fig.savefig('monthly-cumul.png', dpi=300)
fig.savefig('monthly-cumul.svg')

"""
Output is:

count    204.000000
mean       0.827587
std        0.128098
min        0.011139
25%        0.784704
50%        0.857678
75%        0.918354
max        0.985072
dtype: float64

2024-11-30    0.011139
2014-06-30    0.373610
2024-06-30    0.411560
2020-12-31    0.417950
2008-04-30    0.436142
2017-10-31    0.464513
2025-02-28    0.565186
2014-11-30    0.570015
2024-10-31    0.573248
2019-12-31    0.585244
dtype: float64
"""

monthly-cumul.png

monthly-cumul.svg

monthly.png

monthly.svg

pearson.png

pearson.svg