Market timing with moving averages in super

Hi all,

I've always found the 10-month moving average method for market timing popularised by Meb Faber intriguing, particularly in a super environment where you can often switch between options without taxable events. I've run a few simulations using portfoliovisualizer.com on the US share market and AUS total return data and it has been hit or miss since 1980. As expected it tends to underperform in boom markets and outperform in choppy bear markets. This time period is kind of a quirky one though as it includes one of the biggest booms in history (80's and 90's) and the more recent boom (2009+).

Curiosity got the better of me so I downloaded the US share market total real returns data since the 1800's from Robert Shiller's site and ran it through. I ran the analysis on the monthly returns over different periods to see what it looked like, and pretty consistently the timing method beat buy and hold. This was just a simple swap between cash and US share market whenever the 10-month moving average was breached with a first of the month assessment. I've attached the results comparing the market timing and buy and hold results over different rolling time periods from 1931 on. The best periods are typically a little lower for timing, but the average and worst year are often much better. For example for 15yr time frames the average gains were +3% per year and worst years +6% better for the timing method. These numbers are in real returns which is important to me.


I just thought I'd share this as it is worth thinking about for people planning to retire early who have a sizeable super balance. I'm thinking hard about using this strategy for my super investments, and a traditional buy and hold for outside. This is another method of diversification and hence sleep at night support for me. This is definitely not advice, and I'm keen to hear others thoughts on it. Like many things it may already be arbitraged away....