Testing Absolute Momentum on the TSE

A new research paper by Gary Antonacci on absolute momentum piqued my interest.In its simplest form, absolute momentum strategies compare excess asset returns over a pre-defined look back period. If excess returns over the look back period are positive, invest in the asset. If excess returns over the look back period are negative, invest in a 3 month t bill.Antonacci's research shows that absolute momentum strategies work well in a number of markets including US equities, US REITS, US bonds, EAFE, and gold. I thought it would be interesting to see how well an absolute momentum strategy works for the TSE.

For equity data I use the MSCI Canada total return monthly data (includes dividends). For the risk free rate, I use 3 month Canadian t bills. I choose a look back period of 12 months. 12 months seems to work well for other assets so I choose 12 months for my analysis. This minimizes data snooping. The estimation sample covers the period January 1971 to March 2013. For comparison purposes, I also include buy and hold (B&H), a simple MA(10) switching portfolio, and a seasonal switch strategy (invest in the TSE in the 6 months November through April: invest in 3 month t bills for the 6 months May through October). The calculations do not include trading costs.

In the case of Canada, there is some evidence that absolute momentum works. Absolute momentum is preferred to buy and hold because it has a higher Sharpe ratio, Sortino ratio, and Omega ratio. One undesirable feature, however, is that absolute momentum has higher downside risk than buy and hold. Notice how the seasonal switch strategy really stands out. The seasonal switch strategy has the highest Sharpe ratio, Sortino ratio, and Omega ratio. The seasonal switch strategy also has the lowest standard deviatiion and downside risk.