Retail gloom will deepen

Earlier this week I posted an article, The housing-retail link, which discussed the positive feedback loop (“wealth effect”) caused by changes in house prices.

This article argued that changes in housing values are a leading determinant of household consumption expenditure, consumer confidence, employment and growth. That is, when house prices rise (fall) in value, households feel wealthier (poorer) spurring consumer confidence (pessimism), spending (saving) and employment growth (job losses).

I concluded the article by stating that as long as Australian housing values remain stagnant or falling, consumption expenditure, credit growth and job creation will remain subdued, even with Australia’s terms of trade near century highs.

The day after publishing this article, a reader sent me through a spectacular chart confirming the housing “wealth effect” in a very explicit way for New Zealand. This chart, which has been compiled by ANZ Bank’s New Zealand economics team, is reproduced with permission below.

As you can see, the changes in New Zealand real private consumption has tracked the changes in real house prices very closely. And the recent decline in real house prices in New Zealand suggests that the rate of growth of real consumption will continue to fall for the next few quarters.

Further, since the global financial crisis (GFC) hit, and New Zealand house prices began falling in late 2007 (prices are currently down around 15% in real terms since their peak), annual real private consumption growth has been subdued, oscillating between only +2% and -2% over that period. This is a far cry from the heady days of 2002 to 2007, when annual real private consumption growth oscillated between +2% and +7% over that period, fuelled by high levels of home equity withdrawal (see below RBNZ chart).

Clearly, the positive feedback loop “wealth effect” is in play in New Zealand. But what about the other countries referenced in my earlier article: the United Kingdom and Australia?

To gauge whether the same conditions are happening in these countries, I have attempted to re-create the above ANZ Bank consumption vs house prices chart.

First, the chart for the United Kingdom:

Granted, my chart is not as sexy as ANZ’s. But it still reveals a high correlation between changes in real house prices and changes in real household consumption expenditure, thereby suggesting that the wealth effect is in play.

Like New Zealand, since the GFC hit and United Kingdom house prices began falling from September 2007 (prices are currently down around 28% in real terms since their peak), annual real private consumption growth has been in a funk, oscillating between +2% and -5% over that period. Again, this is a far cry from the years prior to the GFC, when annual real private consumption growth oscillated between +2% and +4% over the 2000s, also fuelled by high levels of home equity withdrawal (see below Bank of England chart).

Now consider the chart of consumption vs house prices for Australia:

Like the other countries, the change in Australia’s real household consumption expenditure appears to have tracked the change in real house prices, although the relationship does not appear to be quite as strong as for New Zealand.

With Australian home values performing better in the wake of the GFC, real household consumption has also held-up better than the other countries over this period. However, it remains well below its pre-GFC levels, due in part to households shifting from withdrawing equity from their homes (borrowing) to repaying their mortgage debts (see below RBA chart).

In what is an ominous sign for Australian retailers, the decline in real home prices over the past year has not been fully  reflected in the household consumption figures, which have only just started to turn down from their post-GFC bounce (see above consumption vs house prices chart). If the recent correlation between home values and household consumption holds – and I see no reason why it shouldn’t – then Australian retailers could face several quarters of declining sales growth ahead.

And to make matters worse for retailing, Professor Steve Keen recently showed that mortgage credit growth usually leads house prices by up to six months. So with mortgage credit growth having fallen recently, Australia can expect at least two more quarters of negative house price growth, which should also depress household consumption and retail sales going forward.

The RBA has consistently described the apparent caution in the Australian consumer in terms of a mystery, as if they expect him/her to spring from a shroud and go crazy on the credit card at any moment. Without discounting the possibility that some of that discussion is aimed at jawboning consumers, the RBA may be underestimating its own power. Since 2009, it has explicitly brought house price growth to heel and, by extension, consumption in some degree.

Australian retailers hoping for a return to the heady sales growth of yesteryear might be in for a rude shock. Household consumption and retail sales are likely to remain subdued for the foreseeable future, thanks in no small part to household disleveraging and the sluggish state of the nation’s housing markets.

Cheers Leith

[email protected]

Unconventional Economist


  1. Sandgroper Sceptic

    Awesome charts from NZ there UE. This sort of analysis and the discussion which usually follows is leagues ahead of both mainstream media and financial news sites.

    Hold on to your hats retailers! There are obvious flow on implications for shopping centre owners too.

  2. ceteris paribus

    The correlation of the graph lines is extraordinary. Definitely some sort of causality in there.

    These little graphs alone could be schtik for a world speaking tour on the wealth effect of housing.

  3. ceteris paribus

    PS. With the massive drop of house prices in the US, bread and milk should be the only items moving from the stores there.

  4. Spectacular is the word to describe that NZ chart especially! Off the top of my head, I can’t think of any credible argument for that being merely coincidental.

    While the correlation is blindingly obvious in all of them, it would be interesting to find out why it is stronger in some than in others.

    Are you thinking of looking into this Leith?

    • I suspect it might be due to household savings rates. A higher savings rate should serve as a buffer to falling house prices to some extent, and so reduce the correlation between consumption and house prices.

      NZ has had a significantly lower savings rate than Australia, so they will be far more sensitive to the housing wealth effect. NZ’s savings rate is basically negative over the timeframe of that chart.

      Eyeballing it, the UK chart appears to have a higher correlation between consumption and house prices when their savings ratio has fallen.

      Similarly the Australian chart also appears to show some relation between savings rates and the correlation in consumption and housing prices by eyeballing.

  5. with respect to the retail figures: I recall reading (somewhere) that retiring couples have been buying businesses with their ‘lump sum’ nest egg and that they have been paying 10 and 12 times earnings for these businesses. Historically they attract 3 times.

    What I am suggesting is that yet another entire class of the older demographc is going to hit a rough patch.

    While they won’t be massive numbers it won’t be ‘helpful’ to a troubled economy.

    • “… buying businesses with their ‘lump sum’ nest egg…”

      Indeed Tonydd.

      At this year’s Franchise Expo (Sydney & Perth) the take-up rates and enquires to purchasing new franchises was at an all-time low.

      Some franchises systems in this economic environment we’re heading into just don’t make mathematical sense.

      As one example, you buy a well known coffee franchise for around $600k, need to employ staff and pay exhorbitant rent in large shopping centres with a 5 year (minimum) draconian lease, work 7 days a week with long hours and they sell that dream as ‘freedom to be your own boss’?

      More like indentured servitude in my book.

  6. great job as usual leith. single handedly pulling our housing retail bubble nexus to pieces. i suggest you build your own index ala schiller and get famous. i would like to be analyst for the team when you get it kicked off. haha

  7. Considering the reach your blog has , you should extend this to how it affects government surpluses/defecits as well. Billy has been blogging about this for years.

  8. Bricks and mortar shops are in real trouble. I’ve started buying things directly from China myself recently – things like rechargeble batteries are about a quarter the price they are in the shops. The only things not worth getting directly are the ‘too large or expensive to post’ things. Anyone with lots of exposure to retail and retail property should be really worried – however change is just a natural part of our capitalist system and the bleating of people like Gerry Harvey and Westfield won’t stop it…

  9. Jumping jack flash

    so what will this mean for wage growth and employment. In my opinion it doesn’t look good. how can a business with reduced profits afford to keep staff or raise incomes?

    retail is the bird in the cage, they have direct contact with the consumer and as such are subject to sentiment first. I feel if there is a marked slowdown in retail it is a sign that sentiment has changed and this will soon spread to other sectors.

    • “so what will this mean for wage growth and employment.”

      Bad news when over 60% of our economy is based on consumer spending.

      “retail is the bird in the cage,..”

      Indeed, punters will shut their wallets to discretional spending first. They’ll ‘batten down the hatches’ and try to pay down their debts when sediment changes due to their perceived wealth in assets – cause and effect. The same can be said about the rise in the cost of living.

      Punters are now facing a two-pronged attack!

  10. I like the consumption v house prices graphs.
    I wonder if consumption is including home loan payments?
    Could you tell me the definition used to calulate “consumption”

    If house payments are exluded, then it might be interesting to show also that added to consumption.

    Yours is excellent work.

  11. wealth effect is simply the perception that proceeds greed and fear to drive a market.

  12. This has been obvious for years. Economic growth only has a few sources; the big ones are increases in productivity, increases in population, and increases in borrowing.

    Since 2000, our productivity has gone nowhere. It even fell slightly in some years.

    I believe that the previous government allowed high immigration specifically to inflate GDP growth; the current government deliberately ran very high rates of immigration during the worst of the GFC for the same reason. (GDP per capita fell, but that didn’t get widely reported.)

    That leaves borrowing. Home equity withdrawl seems to be ubiquitous. Many people I know who have mortgages now owe more than they did when they started. When I ask them about this, they say something like “everybody does it,” which is probably true, but that’s a rationalisation for doing it, not a reason. Credit card borrowing is also running at crazy levels; I’ve seen young people with a dozen cards in their wallets. They just don’t seem to care.

    Than again, the fallout from falling real estate prices might not be as dramatic as some people here think. The 3% of our GDP coming in from China will reduce any effects, to some extent. Then again, even that will change once new foreign iron ore production comes on line; that will reduce ore prices dramatically and knock a big hole in the economy that no amount of government policy can fill.

  13. All. I have updated the Australian house prices vs consumption chart by using seasonally adjusted data and shifting the house price data to the right by 6 months. This provided a better correlation, with house prices typically leading consumption.

  14. Seems to me that John Kenneth Galbraith summed up OZ residential real estate in two quotes:

    “In Economics, the majority is always wrong!”

    “Faced with the choice between changing one’s mind and proving there is no need to do so, almost everyone gets busy on the proof.”

    Great work here – keep it up, please

  15. Great article Leith – this type of research has “interesting” implications when articulating an investment in retail stocks – or shorting thereof……

    • Indeed Prince.

      Interesting to see the numbers of Woolworths, Coles, JB HiFi and Harvey Norman when released.

      Another interesting tidbit is Woolworths has started to unload its ownership of shopping centres. What’s that tell us?

  16. Right on. And if Australia’s GDP growth for the March quarter (to be released on June 1) is NEGATIVE (as is likely), the main reason will be the decline in home prices since May last year — not the summer floods, as suggested at .

  17. Great article. Fuel prices must also be having an impact on retail, particularly for people with long commutes. If you are spending an extra $30-$40 a week on getting to work that is less money for other things.

  18. Great work Leith. We all intuitively know that the wealth effect does change our behaviour. I’m not sure if you’ve read much academic research about this effect but there is plenty of good stuff.
    In relation to housing there is this paper which suggests that the wealth effect increases our propensity to consume by 9c per dollar. So if the housing stock of Australia is valued at $3trillion (some say between 3.5 and 4 trillion), and market values increase 10%, then we will spend on average 9% of the $300billion of new ‘wealth’, or $27 billion – with $6 billion of spending occurring in the following quarter.
    There may be studies about the size of this effect in reverse, but if the same values hold in both directions we can look at some interesting scenarios.
    If prices fall 2.5% nationally over a quarter then we lose $75billion of perceived wealth, with an immediate reduction in spending in the following quarter/half year of about $1.5billion and ongoing reductions in spending totalling $7billion
    With about $1.7trillion of bank loans outstanding, that is about the same effect on spending as an increase in interest rates of 0.25% and keeping them there for two years (which will mean $4billion extra is spent on interest repayments per year).
    That’s why house price falls of just a few percent can cascade into a crash so easily. I would suggest the reason that the wealth effect in relation to housing is that many people who benefit/lose from house price changes are highly geared, which increases/decreases their equity more quickly for a given price change.
    On this not I would add that you can’t directly compare share market volatility to house price volatility, since the share market is an equity market. To make a direct comparison you need to compare the volatility of the equity component of the housing market with share market, or the volatility of the share market value plus the value of debts held by those listed businesses to the housing market.

    • While I have disagreed with you on some things in the past, Cameron, that is one extremely helpful posting.

  19. Interesting – the NZ chart shows consumption well in advance of house prices up until 2004. Beyond 04 it’s almost in lockstep. What might be going on here?

  20. For the last 3 years I go to the local Westfields Shopping center for lunch. The last 6months I have seen it turn into a ghost town. And walking around during lunch, if I walk into any store I feel like I am being attacked by shopkeepers eager to see me put my hand in my pocket. It’s insane how bad it is now.
    Although it makes for a better shopping experience 😉

  21. Alex Heyworth

    A convincing demonstration of your thesis since the early 90s, Leith. However, there were plenty of periods earlier in Australia’s history where we had growth and increasing consumer spending without house prices doing anything untoward. The difference this time is easy credit, which I suspect has been driving both house prices and increasing consumer spending. (In the past eras, credit was basically irrelevant to consumer spending, because the availability and social acceptability of consumer credit were both extremely low. Tight credit for housing kept the lid on house prices. Consumer spending was more related to employment than anything else.)

    With the credit tap now tightening somewhat, the downturn in consumer spending as well as house prices is only to be expected.

  22. Some interesting comments re Housing development regulation and land prices, in an OECD Paper:

    Peter Hoeller and David Rae: “Housing Markets and Adjustment in Monetary Union” (OECD, 2007):

    “…..Prudential supervision issues:

    Housing bubbles can be very costly. In Finland, the bursting of the housing bubble in the early 1990s led to a decline of GDP by more than 10% and in Sweden by nearly 5%, even though they eased monetary policy a lot and fiscal policy absorbed part of the shock (Eschenbach and Schuknecht, 2002).
    See Note 11.

    11. In Finland, most of the credit losses came from corporate loans, while credit losses from housing loans were small. But households reacted to excess indebtedness and declining collateral values by accelerating the pay-back of loans, resulting in a rise in the saving ratio. This reduced domestic demand and increased the number of bankruptcies in the non-tradeable sector and thereby the credit losses of banks. The total costs of the banking crisis are estimated at 8 to 10% of GDP. The government provided the banks with around € 16 billion. The net cost to the government was half of this amount (OECD, 2006b)……”