The law of averages

You can use statistics to prove anything

And that's exactly what I intend to do. With reference to some pretty graphs, some interesting statistics and some very dubious reasoning, I lay out the case for why a June pullback is a definite possibility.


After today's strong close the All Ordinaries index rose 3% for the month of May and represents as shown above the 10th consecutive month on month rise in the index. From July 31st 2006 to May 31st 2007 the All Ords has risen an impressive 28.0%. That sounds like quite a stellar performance so I decided to find out how much of an achievement it really is by plotting the frequency of consecutive month on month rises and declines of all durations over the last 10 years (I chose 10 years because that's all the data I have).


The graph above shows that in the past 10 years 10 months of consecutive gains has only happened twice, the first time from May 2004 - Feb 2005 and the second from Aug 2006 - May 2007. In fact it has happened twice in the past 3 years. Another way to look at it is that over the last 36 months the All Ords has risen in 30 of those months. What's to stop the index rising for an 11th, 12th or 18th straight month? Absolutely nothing, but the law of averages suggests we are due for a pullback sooner rather than later.

So that's it, my whole argument basically comes down to the fact that based on historical performance the All Ords is due for a pullback. Not very convincing I know, especially with the economy a picture of health and looming income tax cuts and SMSF perks post June 30th. Oh forget yesterday's soft retail sales number it was just a blip, and ignore the fact that housing is at it's least affordable in years and that major construction companies such as Lend Lease are predicting the housing slump to continue until the second half of 2008. These things interfere with the orderly procession of a bull market.