An Idiot's Guide to PMIs

This subject often divides opinion.

Are PMI's useless or useful?

Let's dive in.

PMI's are considered a leading indicator.

IHS Markit describes their 'Purchasing Managers’ Index' (PMI) as a survey-based economic indicator designed to provide a timely insight into business conditions, providing some of the earliest signals of economic performance.

Great. Let's ask some questions.

What are Purchasing Managers?

And what's so special about them anyway?

Purchasing managers are responsible for making purchases (clue's in the name), which means they receive a lot of information about the firm in order to do their job effectively.

They know when new staff are being hired, they know when production is ramping up or scaling back, they are in the heart of everything.

How do PMI's work?

Keeping it simple, PMI's comprise monthly economic surveys of selected companies.

These are carried out by questionnaire - purchasing managers are asked about changes to output, new orders, employment and prices across key sectors.

Specifically, they are asked if levels are higher, the same, or lower than the previous month.

The responses are then weighted 'to ensure accuracy and representation', adjustments are made for seasonal factors, and this feeds into a diffusion index with a 'score' between 0 & 100.

What's a diffusion index?

A process to summarise the common tendency of a group within a statistical series.

Sounds really boring, right?

Basically, they take a load of inputs (the survey responses) and process them.

If a higher number of the series are rising than declining, the index will be above 50.

If fewer are rising than declining, it will read below 50.

A reading above 50 indicates expansion, and a reading below 50 signals contraction.  

The distance above or below the 50 midpoint represents the rate of change.

Let's look at an example...

It's PMI day.

Last month's PMI reading of 57 represented a solid expansion vs the previous month.

The release is due. Any second now...

52

52 is lower than 57 so that's worse!

Well, it is and it isn't.

A reading of 52 represents continuing expansion/improvement, albeit at a slower rate than the prior month.

In the opposite scenario, a reading of 52 improving to 57 indicates expansion at a faster pace than before.

Likewise, a reading of 42 improving to 47 indicates continued contraction, although the rate of deterioration is slowing.

Labouring the point, but it's a key one.

Reporting of PMI's can be misleading.

For example...

Technically, the headline is accurate, but it can easily be misunderstood, especially if speed-reading.

China's... factory activity... decade high

The word 'growth' is key.

Factory activity is not at decade highs.

Growth in factory activity (vs the prior month) is.

How are PMI's weighed?

According to IHS Markit...

What are Flash PMI's?

These PMI estimates are released a week before the official data, once 85-90% of the data has been processed.

Not all PMI's have a Flash 'preview'.

Why should traders care about PMI's?

PMI's are considered a real-time/leading indicator of economic activity.

Government & Central bank policymakers pay attention to them, so traders and investors should too;

• GDP

• Manufacturing output / service sector output

• CPI/inflation indicators

• Producer input prices

• Exports and trade flow indices

• Employment/non-farm payrolls

• Productivity

• Profitability

• Factory/durable goods orders

• Backlogs of works/capacity indicators

• Supplier delivery times

• Inventory indicators

• Purchasing activity

PMI Weaknesses

One of the main criticisms of PMI's is that they miss out smaller firms.

PMI's provide a broad look at the overall conditions for firms above a certain size (firms who have dedicated purchasing managers),  

Some accuracy is sacrificed to ensure speed of data delivery.

As per the explanation above, diffusion indices are easy to misunderstand and they can throw up some anomalies, especially when the economy isn't working 'as usual'.

Context is key.

We saw this in April 2020...

This sub-index is supposed to represent how well supply is keeping up with demand.

In this instance, a false reading was given (due to the disrupted supply chain) which fed through to the overall reading.

However, the sub-indices for New Orders and Inventories did give an accurate representation of the actual situation.

There is a visual bias to watch out for too.

Exhibit A

There it is folks, right in the middle; a 'V' shaped recovery.

Except it wasn't.

The chart represents the rate of change of economic activity, NOT the overall levels of economic activity.

This is a solid video explainer from the Wall Street Journal;

Now you know the pitfalls, let's look at how PMI's can be useful.

How to read PMI Data

Ignore the headline number.

Know what you're looking for in advance.

The sub-indices are useful, along with the comments from the Purchasing Managers.

Apply context.

If new orders are decreasing while inventories are rising, what does this tell us?

Would that bring the current narrative into question or is it more evidence for continuation?

How could it be interpreted?

Whilst PMI's do not tell the full story, they can be useful to spot potential changes in the wider economic landscape.

Want to go deeper?

The 30 day free trial is back! Sign up here...