Introduction

This algorithm examines the momentum effect in country index exchange-traded funds.

Method

The algorithm picks 35 country index ETFs as the trading universe. As the symbols in the universe don't change over time, we use the MomentumPercent indicator helper method self.MOMP(symbol, period, resolution). This helper method creates a new momentum percent indicator and computes the relative \(n\)-period change in the security. In contrast to the indicator constructor MomentumPercent(period), the helper method indicator will be automatically updated on the given resolution.

In Initialize(), we set the warm-up period to the momentum period and create the dictionary self.data to save the indicator of each Symbol. On each month, the top five index ETFs with the best 6-month momentum will be selected to open long position. ETFs which are no longer in this top list will be liquidated. The Scheduled Events API is used to schedule the portfolio to rebalance at the start of each month.

Conclusion

The algorithm result shows that momentum effects do exist in the country indices. The strategy of holding for one month, a portfolio of the five best performing country indices ETFs over the previous six months, was found to out-perform the equal-weighted portfolio by around 90% per annum over the 22-year period from 2002 to 2022.



Reference

  1. Quantpedia - Momentum Effect in Country Equity Indexes