An update to the popular post on combining TF and seasonality. To recap:

- Trend Following: Price is above 10 month average (per Faber).
- Seasonality: Average upcoming month return (r) over previous y cycles with m periodicity is above a threshold T.
- Dataset used is Fama-French “Small-Value” portfolio from 1954 to 2014.

**AMIBROKER CODE** (commented):

**RESULTS** 1984-2014 (thresholds from 0 to 1%)

**EQUITY CURVE** (T = 0.8):

By increasing the threshold, annual return is almost unchanged but the time in market decreases. For the optimum threshold (0.8%), average monthly return is 2.5% when price is above its 10 month average.

One further improvement that could be made is to normalize by volatility so that the threshold is a function of standard deviation rather than an absolute value. This would allow better testing across instruments.

### Like this:

Like Loading...

*Related*

Why do you use an universe that is not real (Fama/French)? I suggest to use a basket of real stocks to see the differences ….

The reason is access to a 60 years of survivor-free stock data is difficult to obtain. However, the Fama-French stock selection criteria is simple and transparent and closely matched by available ETFs such as VBR.

Hello. Very interesting results. I woulld like to backtest by myself.

But I don’t understand this rule: “Seasonality: Average upcoming month return (r) over previous y cycles with m periodicity is above a threshold T.” can you explain and give an example?

Thank you.

For example, upcoming month is December. Calculate average return, r, for previous n Decembers. If r > T, Buy on Dec 1.

Pingback: The Whole Street’s Daily Wrap for 6/26/2014 | The Whole Street

Thank you. So, in your example, assuming price is above moving average, if 24month average return of the last 15 periods (you cover 30 years) is greater tan 0.8, then you buy?

Hace you tested other periodicities and cycles, for posible curve fitting?

Thank you

Correct.

These are the periodicities I tested: https://rrspstrategy.wordpress.com/2014/06/06/seasonality-roundup-all-timeframes/

Annual (sell in May) is well known but bi-annual is stronger. The autocorrelation drops off very quickly for, say, 23 or 25 month cycles so the effect seems real.