The January Effect

PORTFOLIO POINT: The ASX’s strong start to 2012 makes it worth examining whether the January Effect is at work, and what it might mean for the year.

The “January Effect”, if you don't already know, is a market theory. And with our ASX rising more than 5% in January 2012, it's a theory with special relevance this year. The notion that stocks are more likely to rise in the first month of the year goes back to 1942, when it was first advanced by Wall Street banker Sidney Wachtel.

He put forward some evidence that US stocks moved higher in the first month of the calendar year, and specifically that small stocks outpaced larger stocks during the period.

After several decades in which the January Effect has become a more widely known theory, it is now also used as a predictor. So what does it mean for Australian investors this year?

Obviously, we are unlikely to see every month perform as well as January 2012 or we would be heading for a return of better than 60% for the year, and that is very unlikely.

But let's look at the numbers: The following table looks at the January returns and then the total year returns since 1990. A quick look shows that not only does the January Effect not seem to be predictive of the year, it almost seems to be a contrarian indicator, being right less than half the time. (As an aside, at the end of last year we heard a lot of commentary that markets had provided two years of negative returns. In 2010, sharemarket values fell but including dividends made the overall return slightly positive.)

-Has January been a fortune teller?
Year	January return	All Ords returns (inc dividends)	January Effect predictive
1990	1.60%	-17.50%	NO
1991	3.30%	34.20%	YES
1992	-1.70%	0.50%	NO
1993	-1.80%	40.50%	NO
1994	6.30%	-8.20%	NO
1995	-3.90%	20.20%	NO
1996	3.30%	14.00%	YES
1997	0.10%	11.40%	YES
1998	1.20%	8.50%	YES
1999	2.00%	19.30%	YES
2000	-0.70%	5.00%	NO
2001	4.30%	10.10%	YES
2002	1.35%	-8.10%	NO
2003	-1.40%	15.90%	NO
2004	-0.70%	27.60%	NO
2005	1.28%	21.10%	YES
2006	3.50%	25.00%	YES
2007	2.00%	18.00%	YES
2008	-11.00%	-43.20%	YES
2009	-5.10%	37.00%	NO
2010	-8%	1.60%	NO
2011	0.20%	-10.50%	NO
2012	5.10%		?

But we need to look long term. AMP analysed the 109 years of Australian sharemarket returns from 1900 to 2008 and found that in 21 years (or 19% of the time) sharemarket returns were negative, and in 88 years (81% of the time) they were positive. The AMP data provides an interesting perspective on how challenging times have been: 2008 (a market return of –43%) is the worst year for returns in the past 108; followed by 1930, with a return of –30%.

We often talk about the 1987 crash as being a tough time, but the return for calendar 1987 was –7.9% (prior to the October crash sharemarkets has risen significantly), and then following 1987 only two out of the next 14 years provided negative returns.

In the 22 years of sharemarket returns we have put together in the table – capturing the early 1990s’ recession, the Asian currency crisis, the dotcom bubble bursting, September 11 and the global financial crisis – the January Effect has predicted what has followed in 10 of the 22 years.

This is a success rate of just 45%. We know that in the majority of years sharemarkets provide positive returns. In our sample total, sharemarket returns (capital growth and dividends) were positive in 15 of the 22 years, or 68% of the time. It is interesting to note that this is less than the historical long-run result calculated by AMP, of positive returns 81% of the time, reflecting the challenging time that investors have had.

This data suggests that there is little value in following the January Effect as a possible source of information about future annual sharemarket returns. Guessing that sharemarkets will go up in any year will make you look like a genius 68% of the time based on the data from the past 22 years, whereas going with the January Effect as a forecasting tool had the somewhat embarrassing record of predicting the annual outcome less than 50% of the time: worse than tossing a coin or simply guessing the market would go up.

However, as noted earlier, January this year has been an extraordinary month, up 5%, and we investors are entitled to a closer examination of the theory. To be honest, I wanted to interrogate this data until it told at least some sort of optimistic story.

So with my integrity temporarily suspended, I tried the hypothesis that when January returns are significant moves, it becomes predictive of the year’s returns. In nine of the past 22 years, sharemarkets were either up or down (a significant move) by more than 3%. In these years it had predicted the returns from the sharemarket five times – a success rate of 56% of the time

Hmmm...not bad. Time for a little more data mining.

Changing the definition of a “significantly positive January return” to years in which shares rose by 2% in January, brings up seven of the 22 years. Of these, subsequent returns were positive in six of the years, a success rate of 86%. Let's round this out to 90% (what's a bit of rounding between friends).

But wait '¦ there's more. In the 90% (allowing for some rounding) of years when shares provided positive returns following a “significantly positive” return for January, the average return was 20.1%.

So there we have our theory.

Based on the past 22 years, when January provides a significantly positive return, 90% of the time shares have provided a positive annual return of more than 20%!

With shares up 5% for January this year it is time to get the credit card out and lever up. After all, what could possibly go wrong?

Scott Francis is an independent financial adviser based in Brisbane.