Original paper

SSRN-id329840.pdf470.1KB

Abstract

Stock market price/earnings ratios should be influenced by demography. Since demography is predictable, stock returns should be as well. We provide a simple stochastic OLG model with a cyclical structure which generates cyclical P/E ratios. We calibrate the model to roughly fit the cyclical features of historical P/E ratios.

Keywords: Demography, Price Earnings Ratio, Returns, Efficient Markets, Baby-boom, Savings

Trading rules

Target investment: Focus on 15 developed countries (Australia, Austria, Belgium, Canada, Denmark, France, Germany, Italy, Japan, Netherlands, Spain, Sweden, Switzerland, UK, US).
Evaluate the proportion of adults over 65 in every nation.
Determine the mean and variability for this proportion.
Adopt a long position in nations where the elderly (population >65) constitute a percentage that's at least 1 standard deviation under the mean.
Adopt a short position in nations where the elderly represent a percentage that's at least 1 standard deviation over the mean.
Adjust portfolio positions once every year.

Python code

Backtrader

import backtrader as bt
import pandas as pd

class DemographicStrategy(bt.Strategy):
    def __init__(self):
        self.country_data = pd.DataFrame({
            'country': ['Australia', 'Austria', 'Belgium', 'Canada', 'Denmark', 'France', 'Germany', 'Italy', 'Japan', 'Netherlands', 'Spain', 'Sweden', 'Switzerland', 'UK', 'US'],
            'percent_over_65': [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]  # Fill in actual data
        })

    def next(self):
        if self.datetime.date().month != 12:
            return
        mean_over_65 = self.country_data['percent_over_65'].mean()
        std_over_65 = self.country_data['percent_over_65'].std()
        long_countries = self.country_data[self.country_data['percent_over_65'] < mean_over_65 - std_over_65]
        short_countries = self.country_data[self.country_data['percent_over_65'] > mean_over_65 + std_over_65]
        # Allocate equal weights to long and short positions
        long_weight = 1.0 / len(long_countries)
        short_weight = -1.0 / len(short_countries)
        # Adjust positions
        for d in self.getdatanames():
            country = d.split('.')[0]
            if country in long_countries['country'].values:
                self.order_target_percent(data=d, target=long_weight)
            elif country in short_countries['country'].values:
                self.order_target_percent(data=d, target=short_weight)
            else:
                self.order_target_percent(data=d, target=0.0)

cerebro = bt.Cerebro()
# Add data feeds for each country's ETF
# Data should be in a format that includes the country name in the dataname field
# Example: cerebro.adddata(bt.feeds.YahooFinanceData(dataname='Australia.ETF', ...))
# ...
cerebro.addstrategy(DemographicStrategy)
cerebro.run()

Please note that you will need to replace the ‘percent_over_65’ values in the ‘country_data’ DataFrame with the actual data. Also, make sure to add the data feeds for each country’s ETF to the ‘cerebro’ object before running the strategy.

Demography and the Long-Run Predictability of the Stock Market

Original paper

Abstract

Trading rules

Python code

Backtrader