Following my post last week Let Retail Burn, Cameron Murray has followed up with a series of log-scale charts that show real retail sales growth per capita since the early nineteen eighties. Firstly, Cameron offers the overall sales picture (I have added trend lines in red):
As you can see, there is an accelerating trend that really took off in the mid nineties and peaked in late 2004. That tracks the Australian credit bubble pretty nicely to my mind.
Cameron goes further with a second post that breaks up the internals of sales into segments:
My approach is to examine retail from a household perspective. Rather than look at total turnover in current prices, I will examine real spend per capita in each of the main retailing subsectors. I do this because economic theory has a lot to say about changes to household spending patterns during economic cycles.
Economic theory would suggest that in boom times, retailers of luxury goods would see turnover increase more rapidly than incomes. As Wikipedia explains – In economics, a luxury good is a good for which demand increases more than proportionally as incomerises. The reverse should also be true for these goods.
Importantly, retail trends need to be seen in the context of a housing driven wealth effect. The wealth effect is an increase in spending that accompanies and increase in perceived wealth, rather than spending which is driven by growth in incomes.
The wealth effect is also behind many of the saving decisions of households. Since 2005 the trend of declining household savings rates was dramatically reversed. We now have a household saving ratio not seen since 1987 (see the RBA’s chart below). This is an important backdrop to the retail story.
These factors are important to consider if you foresee near term home price declines. In this scenario, spending in wealth driven retail sectors would be expected to fall more than flat or falling household incomes, and increased savings alone would suggest.
Now to the detail.
The graphs below show the performance key subsectors in retailing. Note the log scales, which mean a straight line indicates a constant rate of growth – the steeper the line, the higher the rate of growth. Note also that this is a real per capita measure, which is indicative of trends in household spending decisions. Quarterly chain volume data is used, with May 2011 current price data adjusted to substitute for June 2011 data. The ABS explains some of the trends in more specific subcategories here (definitely worth reading the context of this post).
A few points jump out at me from the graphs. First, household goods (maroon in first graph) have outperformed by a long way, for a long time. This category includes furniture and appliances, hardware and gardening, floor coverings and electrical. This sector also appears to have seen the sharpest shock around the end of 2007 – from having the strongest rate of growth to nearly the weakest. The rising part of the curve might partly be attributed to a greater appetite for expensive furniture and appliances, which is indicative of a luxury good effect. Also important is the impact of the construction boom of the early 2000s which has since collapsed in many areas.
Second, clothing and accessories (green line) was on a declining trend for 14 years until 1997. For a decade since then, the growth rate in this sector was only bettered by household goods. Spending recovered strongly since the GFC. I’m not exactly sure why this might be the case. Perhaps some readers have experience in this sector.
Food retailing has been the steadiest (as you would expect) with only a slight easing from the growth trend since 2009 (maroon in second graph).
Other retailing (which includes pharmaceuticals, recreational goods, cosmetics and books) appears very sensitive to the housing wealth effect, seeing big spending boost during the 2002-03, the 2007, and the 2009 house price booms. Surprisingly spending has remained strong since the GFC – the only retail sector where this has occurred.
We might attribute some of the recent robustness to the high Aussie dollar. The ABS explains that pharmaceuticals and cosmetics and toiletries are the strongest components of this sector.
Cafe and restaurant spending (orange line) also appears sensitive to the wealth effect, and is noticeably one of the more volatile sectors.
Department store spending has been declining steadily since the end of 2007 (purple line). Anyone who had closely examined this data would not have been so surprised about David Jones’ recent profit downgrade. Spending at department stores is now back where it was in 2003 on a per capita basis.
Finally, the second graph has the period of 2002-03 circled. This is simply to highlight that all retail sectors grew at abnormally high rates during the house price boom of this period. Indeed, we can see the wealth effect correlation between house prices and retail growth in many sectors in 2007 and 2009, although to a lesser extent.
My near term outlook is for a subdued retail sector. As I have said before, I believe that in these challenging times for retailers, innovation will be the key to staying ahead. New business models that use internet shopping to good effect, with a small physical store presence might be one path for many. Those companies who adapt quickest will benefit.
Could not have put it better myself. Thanks Cameron.