5/1/17: Gwan Ya Beaut... Irish PMIs ≠ Irish GDP

Some years ago, I have shown that Irish measures of economic activity - when collected at sectoral levels - have virtually nothing in common with Irish GDP and GNP. Given recent revisions to economic growth and the National Accounts, including the absurd levels of notional GDP and GNP growth recorded in 2015 and in parts of 2016, it is worth to revisit the same issue.

So here is the data: the best advanced indicator data on Irish economic activity that we have is the set of Purchasing Managers Indices (PMIs) released by Markit for three key sectors of the economy: Construction, Manufacturing and Services. Markit are doing pretty much an honest job surveying companies to determine if they are experiencing uptick or decline in their activities. And they are doing this every month. Yes, there are issues with data quality due to what appears to be a strong pro-MNCs bias in the surveys. And yes, Markit are refusing to fully investigate the matter and to test data formally for such biases. And yes, Markit are still not willing to share with me their data, including the actual final data set of PMIs (I have to collect these manually, every month).

But, for all the above problems, Markit is the only source of leading economic indicators for Ireland.

So next is the question: do rates of growth signalled by PMIs actually relate to the rates of growth recorded in the economy (GDP and GNP)?

Let’s take a look, using CSO’s official National Accounts data.

The above shows whatever is happening in Manufacturing. Nope, growth rates signalled by PMIs are not correlated with growth rates in GDP or GNP.  Changes in Manufacturing PMI signals account for only 9.3% of variation in GNP and 6.4% variation in GDP. You wouldn’t be asking Manufacturing sector for its view if you wanted to gauge Irish aggregate economy. 

The above shows what is happening in Services. Again, growth rates signalled by Services PMIs are not correlated with growth rates in GDP or GNP.  Changes in Services PMI signals account for only 12.6% of variation in GNP and just under 8% variation in GDP. You wouldn’t be asking Services sector for its view if you wanted to gauge Irish aggregate economy either.

Why are both sectors signals come out utterly useless when it comes to signalling growth in either GDP or GNP? We have no idea. But my speculative view is that in reality, even large MNCs can’t organically establish their own ‘contributions’ to Irish GDP because whilst purchasing managers and related executives on operations side might know what their divisions are doing and how much more or less business they are handling, the same managers have no idea what value in the end will be attached to their divisions work by the finance lads on the Mother Ship. In other words, real operations managers have no clue how much their companies are booking in revenues or profits because these revenues and profits have only tangential connection of Irish operations. Tax arbitrage is such a naughty thingy, you see, when it comes to collecting data.

Not that Markit (or a vast array of Irish stuff brokers so keen on using its data to ‘interpret’ ‘buy everything’ signals for Irish assets) mind… Gwan, ya beaut... buy some stocks, will ya?