It’s no secret that I am a big fan of simple asset allocation strategies. They are able to combine quantitative market analysis with trading strategies that are simple enough to explain to anyone. Strategies like the Dual Momentum ETF Strategy または、 Ivy Ten Portfolio offer a simple way for traders to take a trend following approach to global markets in their own personal accounts.
One thing that I have always wondered about these types of strategies is how their performance would be altered by changing the day of the month that you rebalance them. Would the results be dramatically different if you were to rebalance the portfolio on the 15th of each month instead of the first?
The authors from Flirting with Models had the same question when they wrote their post about The Luck of the Rebalance Timing. その記事で, they broke down one of Mebane Faber’s popular asset allocation strategies and backtested the strategy using different days as the rebalance date.
10-Month Simple Moving Average Timing Model
The strategy that they used for that article was Faber’s 10-Month Simple Moving Average Timing Model. This strategy trades a basket of ETFs and uses whether they are above or below their 10-month simple moving average to determine if they should be in the current portfolio. しかし, in an effort to simplify everything, they focus exclusively on trading SPY.
The results showed that changing the rebalance date by one day can actually make a significant impact. この場合, using the 19th trading day of the month resulted in annualized return of 11.33% 最大ドローダウンと 19.03%. Changing the rebalance date to the 20th trading day of the month adjusted those results to an annualized return of 10.51% 最大ドローダウン 30.22%.
The Ivy Portfolio
The authors recently published a follow-up article that performed the same type of test on Faber’s Ivy Ten Portfolio. もう 1 回お願いします, the returns varied a great deal depending on what day of the month was used to rebalance the portfolio.
この場合, the best portfolio would have returned 9.93% で 2013. The worst portfolio would have returned only 4.22% that year.
How To Account For Luck
明らかに, there is no way to determine which day of the month will be the best to rebalance a given portfolio in the future.
The authors explained that the best solution would be to divide the capital into 21 different portfolios that would each rebalance on a different day of the month. This would result in the overall portfolio becoming an average of all of the possible rebalance days. This assumes that you have adequate capital, plenty of time, and transaction costs are not an issue.
The authors recognize that rebalancing on a daily basis defeats the purpose of these types of strategies, ので、 the solution that they propose is to divide the portfolio into four portfolios and rebalancing one of them each week. Their results show that this smooths out much of the rebalancing risk while still not requring a large amount of time or transaction costs.