Wednesday 24 October 2018 by Opinion

Predicting market movements in the new world order – Part 1

Craig Swanger explains how he assesses markets why he thinks RBA cash rate rises are dead in the water.

Prediciting Movements Article

Two years ago we had clear signs of:

  • Chronic underemployment and an overheating housing market in Australia
  • Trump threatening trade wars
  • The Fed clearly on a tightening path well ahead of the rest of the world
  • Overvalued equities markets
  • Heavily indebted China
  • The EU struggling with Italian debt and populism

Everything we needed to know to predict today’s market was available this time last year, much earlier in many cases.  Yet it wasn’t until markets actually fell, that bank economists started predicting more falls. 

Why economists keep getting it wrong

Like the old saying goes, you can’t see what is in front of you by looking backwards.  Yet economists fill their days doing just that.  Their blind use of models suggests that there is some law of nature that says that economic forces must act tomorrow as they did yesterday.  That is lazy at best, dangerous at worst. 

Right now more than at any time for more than 60 years, models based on the past will fail to predict the future.  The world economy is changing permanently.  This isn’t new; there have been major revolutions every few generations.  When the Industrial Revolution changed the world forever, it broke all of the economic models of the time.  Countries with large pools of capital could suddenly produce more than countries with large pools of labour. 

Fast forward 150 years and the digital economy has lowered prices permanently, shifted labour patterns, and increased social transparency, which impacts respectively:

  • The Philips Curve that once helped predict rising inflation as unemployment fell, has been rendered not just useless, but dangerous
  • The slow death of unemployment as a useful measure of economic health
  • The rise of social power finally able to demand transparency from corporate giants (witness the recent fates of Facebook, Australian banks, and our aged care sector) and politicians.

These trends were painfully clear if you are looking outside and forwards, not at your computer and at yesterday’s models.

So where should we be looking?

Predicting the future is never easy, but intuition often provides the first clues.  Data can then be used to test whether there are early signs that your intuition was right.  Because investing in bonds is a long term strategy, not the domain of day traders, I have always looked at trends that could drive prices over the next 3-10 years.  Using some of those predictions as examples of where to look for tomorrow’s market trends shows where some clues might be hiding.  Below and then continued next week, we give three examples of where the largest trends to hit investment markets in recent years offered us a lot of clues a long way ahead of their impacts of markets:

  1. The fall of Australian government 10 year bond yields below the US equivalent
  2. The rise of populism (again)
  3. China forced to borrow more

​​1. Australian 10 year bond yields fall below US 10 year bond yields

This prediction was first made late in 2016 when Australian 10 year bonds were yielding 2.77%pa and US rates were 2.133%.  I first made the comment at a FIIG client seminar and it was certainly seen as controversial, particularly given literally not one economist agreed.  Eighteen months later and the Australian 10 year yield fell below the US equivalent for the first time since 2000, but even then bank economists claimed “it wouldn’t endure”.  Still looking backwards to find what was in front of them. 

The key was record high household debt likely to mean the RBA would need to see significant inflation risks before risking economic stability by rising rates, and extremely weak inflation being caused by the digital economy and rising underemployment.  The US on the other hand had managed to get its underemployment down from record highs and was facing tight labour markets and extraordinary stimulus from Trump’s aggressive company tax and trade policies.   

This information was available and intuitive.  Labour markets around the world were changing and while more people had jobs, most people were working less. 

In Australia, ABS labour market data showed the longest stretch of underutilised labour (unemployed + people working less hours than they wanted to) above 13% since the 1990s recession.  Yet the headlines kept celebrating low unemployment and linking that to the likelihood of rising wages and inflation. 

 Year after year the headlines were wrong, including the RBA’s own forecasts for a staggering seven years in a row.  Wage growth still eludes Australia, but not the US.  The difference is underemployment. 

Figure 1 clearly shows the worse that Australia’s underemployment gets compared to the US, the lower the interest rate differential.  The blue line is the difference in underemployment, with inverse scale so that the lower the blue line, the better the US employment situation relative to Australia. 

In 2014, the US economic improvement showed clear signs of outpacing that in Australia as the blue line on the right hand chart fell further and broke away from the red line (interest rate differential).  Eventually, the interest rate differential needed to rebalance, or we’d need to believe that the relationship was broken.  As wage growth in the US was rising but not in Australia, it lead us to believe back in 2014 that rates in the US would outpace Australia, pushing down the AUDUSD exchange rate.  Then when the gap worsened for Australia by 2016, we predicted the 10 year bond rates would cross. 

Underemployment and long term bond yields: US/Australia differentials

 Predicting movements chart 1

Source: ABS/ Bloomberg
Figure 1

So there is the first lesson for the new world order: Unemployment won’t drive interest rates, “underemployment” will.

 The economists’ favourite tool, “The Phillips Curve”, tells them that low unemployment = rising inflation.  The theory is good; if there is less labour available, employers have to pay more to attract new employees.  But using unemployment as a measure of labour availability doesn’t suit this modern era called the “Gig Economy” in which so many people work part time and some work full time but in two jobs.  The Phillips Curve needs to be modernised to tell us that low underemployment = rising inflation. 

 So next time you see the headline “Record low unemployment to see interest rates rise”, ask yourself what is the underemployment rate doing?

 Related to this issue is the rise of populism, which I’ll cover next week.  As a segue worth keeping an eye on though, have a look at the underemployment rate amongst Australia’s 15-24 (“youth”) market. 

The Figure 2 illustrates the extent of this worrying trend in Australia.  It shows the long term unemployment rate which in the case of both men and women is rising slowly, showing a structural rather than a cyclical shift.  Furthermore, the rate for men is still well below the 1990s recession but for women it is rising steeply towards an all time high, suggesting the impact is being felt in sectors more dominated by women, like retail. 

 Australian long term unemployment rates

 Predicting movements chart 2


Source: ABS
Figure 2

At the same time, there is increasing underemployment despite lower unemployment.  More people are working, but not getting the work they want. 

The unavoidable conclusion is that there is significant structural “slack” in Australia’s labour markets, and until that is addressed, wage growth will be weak.  This means inflation in Australia will be weak, which in turn means interest rates won’t rise, and certainly won’t rise as fast as the economists’ Phillips Curve would tell us.

Next week: Linking underemployment to the rise of populist politicians and a rising threat to the Chinese economic miracle