The January effect

Ever heard of the January effect? Well you're bound to in the coming weeks . it's a theory which emphasises the importance of this first month of the calendar year to sharemarket investors.

Why? Because many people believe that January sets the tone for the markets, especially on Wall Street.

You'll also find regular references to a notion that January, due to the tax dealings of US investors, is a more profitable month than any other.

Is any of it true? We can find out by asking ourselves two questions.

1. Are January returns predictive of market returns for the year?

2. Does the USA January effect have relevance in Australia?

From an investor's perspective the predictive effect of January is as appealing thesis. If returns in January forecast returns for the year ahead then perhaps that can applied in a profitable strategy. An investor following this process might significantly overweight shares over the course of years where returns are positive in January and significantly underweight them in years where returns are negative in January.

But anyone with more than a passing interest in markets will know just how poorly the 'January effect' predicted market moves in 2009. After the index fell 5.1% in January it recovered strongly to deliver a return of more than 35% (total returns - including dividends and capital growth).

Still, one year can hardly serve as a signal - looking further back we know that share markets provide negative returns in approximately three out of every 10 years. Therefore, if at the start of every year we were just to guess that share markets would go up, we would be correct 70% of the time.

On that basis the January effect has to show a greater predictive ability that being correct 70% of the time, which we can do by simply guessing that markets are going up all the time, to be valuable.

The following table considers the market timing January effect (using returns for the All Ordinaries Accumulation Index) for the past 20 years.

The second column shows the return from January, and the third column the annual return in that year. The fourth column then comments as to whether or not the January effect was predictive of the full year returns.

The bottom line was that over the past year the market timing January effect predicted returns correctly in only 10 out of the 20 years. That's half the time, but from an academic perspective this is hardly predictive at all: That's because just guessing that the market would go up each year correctly ?predicted' movements in 15 out of the 20 years.

Separately, there was no evidence that the market timing January effect seemed to do better in the year's that really mattered, that is the years when the market either roared or slumped.

While it did predict the 43% downturn in 2008, that was the only negative year that it predicted correctly. Negative January's in 2009, 1993 and 2004 would have seen investors miss out on outstanding returns of 37%, 35% and 22% in those years.

These were three of the best four years over the past 20 years, and it goes without saying that investors could not expect to receive a reasonable investment return over the past 20 years while missing out on well above average returns in three years.

So much for the predictive effect of January ... it's just a myth.

But what about the notion that January itself is a great month to invest. Here at least there is some evidence. In the US, stockmarkets shares tend to have higher performance over the month of January, with small company shares doing particularly well.

We know from a range of studies over a range of periods that the average long run return from US sharemarkets has been around 12% a year. This translates to a return of 1% a month.

Now the January effect as described in US share markets has been for an 'excess return' of around 1% for January - that is, returns for January have tended to average 2% a month rather than the more usual 1%. So to be fair, there is some truth in the theory that returns are effectively double in January in the US. Of course we only talking about the difference between a 1% return and 2% return, hardly something the average Australian investor could profit from.

What really happens in the US in January is the American investors often sell shares that have made a loss toward the end of their tax year (December), and then buy them back over January. This leads to higher demand for shares, particularly in small companies which individual investors might tend to overweight, pushing prices up.

In Australia our June end financial year means that any tax related selling in our market would be at that time of year. Indeed, there has been some evidence in some academic work that we have an 'August effect'. Certainly, recent warnings from the Australian Tax Office have advised investors to be careful around tax related trading at that time of the year

The two January effects that we looked at are both interesting - although don't seem to lead to investing strategies that will provide Australian investors with above average returns.

Indeed, any investor who had followed the January effect with the intention of using it to 'time the market' would have found that strategy costing them investment returns - understanding that the strategy is not predictive will help you avoid those mistakes in the future.