Argentina is the quiet achiever in global dairy industry trade.  They keep ticking along at a growth rate of about 2.5 – 3.0% and every now and then they put in a spurt. This year they are having a real crack. The chart below shows the monthly milk production for the past 7 years and our seasonally adjusted plot. The seasonal adjustment shows the extent to which milk production is ahead of or behind the long term trend line.  The percentage growth is calculated relative to this long term trend. It is not biased by unusually high or low milk production in the year prior.

The chart shows that monthly milk production is close to 100 million litres above the long term trend. In this calendar year it looks like they will grow their industry by more than 10% or over 1 billion litres. In global terms Argentina has gone well past Australia as a milk producer. At this rate of growth they will surpass Australia’s dairy exports in 2 – 3 years.

For those of you that just want know what the milk production trends for Argentina are that’s all I have. You can stop reading now. The real aficionados of this blog (all 3 of you) may however be interested in an interesting diversion into numerical modelling.  If so read on …

What really attracted my attention about Argentina this month was the release of their official data for total milk production from January – June 2011.  Unlike our New Zealand neighbours (you really have to drag it out of them), the Argentinian Ministry of Agriculture provides excellent and timely release of monthly milk production data. The monthly data is however an “indicator” of milk production. It is the volume received by a selected group of processors. It represents about 65% of total milk production. There is a consistent relationship between the indicator volumes and actual and so it is possible to estimate total milk production from these figures.

Xcheque’s monthly milk production data for Argentina is based on the indicator estimate but we update with actual when this becomes available (6 – 9 months later). This official total data gets me excited because I can see how well the estimated figure is tracking against the actual. You can therefore imagine my consternation when the actual data for Jan – Jun 2011 came in at about 2% above our estimate. That is 100 million litres extra on the world market! … “We’ll all be rooned”, said Hanrahan.

A quick check of the data and our model brings a sigh of relief. Yes the recent actual data is tracking above the model estimate but that is just part of the normal error in the model. Now with just a small adjustment to the long term bias we can close that variance down to 25 million litres – just a round of (milk) drinks in the context of both Argentina and global dairy production … and while we are at it, let’s adjust the seasonal production calculation for trend bias as well.

The chart below shows the actual versus our estimate from 2007 – 2011. We like it, and our thanks go to the Argentina Ministry of Agriculture for providing the nice data.