Commentators misunderstand how interest rate changes affect retail sales growth

Charlie Nelson
January 2012

After saying for most of 2011 that they would have to raise interest rates, the Reserve Bank of Australia (RBA) cut interest rates by 0.25% in each of November and December 2011.  While this is a tacit admission by the RBA that they misread the economy, the rate cuts have been greeted by commentary that completely misunderstands how interest rates influence consumer discretionary spending (as measured by retail sales).

The Melbourne Cup day interest rate cut has failed to spark the shopping spree retailers were hoping for, with spending growth stalling in November.  Despite interest rates falling by 25 basis points during the month, the first official figures on the early weeks of Christmas shopping suggest the festive season began with a whimper.
Clancy Yeates, The Age, January 10 2012.

Such an initial reaction to an interest rate change is consistently observed.  This is because there are some people who like higher interest rates – net receivers of interest – who receive less income when interest rates fall and so spend less.  Those who have debt react in the opposite direction but usually much more slowly and it takes up to 18 months before a change in interest rates has its full impact on consumer discretionary spending.  See www.foreseechange.com/perverse.pdf for an explanation of this written by me about 10 years ago.  The RBA denied any perverse reaction in their 1995 annual report (page 35) and have clearly never understood how different consumers react.

Economists, economic commentators, retailers, and their industry associations obviously do not understand either.

The Reserve Bank of Australia may need to cut interest rates further if it wants to jolt consumers out of their pessimistic mood.  There has been little improvement in household confidence despite two interest rate cuts.  The Westpac-Melbourne Institute consumer sentiment index rose just 2.4% this month to 97.1 points.
Adrian Rollins, Economics correspondent, The Australian Financial Review, 19 January 2012.

This is doubly misleading commentary because not only has Rollins displayed no appreciation of the lags but also he does not acknowledge that the index of consumer sentiment has little or no correlation with retail sales growth – see www.prophetsprofit.com.au/misleadingics.htm.

Interest rate cuts have done little to boost Australia’s retail trade figures, a retail industry group says.  Australian National Retailers Association (ANRA) chief executive Margy Osmond has labelled the November 2011 retail trade figures “disappointing” and says a third rate cut in February could be the best medicine to boost the floundering industry.  “Certainly we would have thought in the past that two interest rates cuts would have been enough to stimulate spending – well maybe it takes three now,” Ms Osmond told reporters in Sydney on Monday.
Australian Financial Review, 9 January 2012.

Interest rate changes take considerable time, many months, to influence the economy more generally, and yet ...

The Melbourne Cup day 25 basis-point shave would only just be starting to ripple through the economy and it probably won’t be clearly evident until later this month whether it had the desired effect.
Ian Verrender, The Age, December 6, 2012.

Researchers at the Reserve Bank of Australia published an analysis of the lags of monetary policy in 1997 (RDP 9702).  They found that the average lag of about five or six quarters in monetary policy’s effect on output growth.  There is, of course, considerable uncertainty about predictions of the lag for any particular episode.  Interestingly, in that paper the researchers identified six main channels through which changes in interest rates affect economic activity.  One of these was “cash-flow effects on liquidity constrained borrowers”.  They did not, however, mention cash flow effects of net recipients of interest (the perverse effect).  They have implicitly assumed that there is an “average” consumer with the same lagged response to interest rate changes.  There is no such animal!

Chart 1 shows interest rates (banks standard variable mortgage rate) and retail sales growth over the past two decades.  There is no consistent contemporaneous correlation.

Interest rates were rising between mid-2002 and mid-2008 and yet retail sales growth went through ups and downs and peaked in 2007 just before interest rates peaked.  Interest rates were falling between late 1990 and late 1993 and yet retail sales growth went through ups and downs and then peaked between the June 1996 and March 1997 quarters as interest rates peaked.

Chart 1

That there is no consistent relationship between interest rates and retail sales growth is because:

  • Lags are long and variable.  Sometimes there is a quick impact but mostly this is in the opposite direction expected by most commentators – this is the perverse impact by net recipients of interest.
  • The direction of impact is both positive and negative.  Net recipients of interest tend to react quickly while net payers of interest tend to react slowly.  The slow reaction of net payers of interest is because many of these borrowers are paying off more than the minimum required amount on their mortgage so it may take several rate rises before they are required to make higher payments.  When interest rates are falling, many people with a mortgage do not pay less – they prefer to keep paying the same and reduce the term of their loan.
  • There can be threshold effects.  This was observed in early 2008, before the global financial crisis hit.  The last rate rise in late 2007 and the further two rises in early 2008 had a disproportionate effect compared with the many earlier rises.  This is discussed in detail below.
  • Other factors influence retail sales growth, including house prices, the share market, and motor fuel prices.  Without quantifying the impacts of all these as well as interest rates, the impact of interest rates cannot be reliably discerned.

Interest rate preferences have been tracked by foreseechange since 2005 (Chart 2).  The surveys are administered by telephone, are of a nationally representative sample, and the sample size each wave was 500 increasing to 1,200 since 2010.

Chart 2

 

On average, 39.7% of adults prefer interest rates to go down; 17% prefer interest rates to rise; 39.1% prefer no change; and 4.0% don’t know their preference.

There was a marked change in interest rate preferences between October 2007 and April 2008.  The RBA raised interest rates by 0.25% in each of November 2007, February 2008, and March 2008.  This had the effect of increasing the proportion which preferred lower interest rates from 37.5% in October 2007 to 56.3% in April 2008.  The proportion which preferred no change fell from 44.8% to 28.8%.  Clearly, many people who earlier thought that higher interest rates were not hurting suddenly changed their mind.  Retail sales growth, which was grew by 8.0% in the December 2007 quarter (compared with the December 2006 quarter) plummeted to a low of 3.1% in the September 2008 quarter.

This was neither a prescient nor a wise piece of engineering by the RBA on the eve of the global financial crisis.

At the same time, between December 2007 and June 2008 quarters, the proportion of household disposable income being used just to pay interest shot above 12% (Chart 3) and this must be currently a critical threshold above which the reaction of consumers to interest rate changes is non-linear.  The threshold seems to have been about 8% on the eve of the recession of the early 1990’s.

By the March quarter of 2009, home loan interest rates had fallen to 5.85% from a peak of 9.6% in August 2008.  Over this period, interest payments fell from 13.5% of household disposable in come to 10.0%.  By April 2009, the proportion of adults preferring lower interest rates had fallen back to less than 40% and the proportion which preferred no change had increased significantly.  Retail sales growth increased significantly over this same period.

Chart 3

 

Interest rate preferences by age group for 2010 are shown in Chart 4.  While there are about 10% in each age group under the age of 50 who prefer higher interest rates, preference for higher interest rates increase over the age of 50 and over the age of 60, more prefer higher interest rates than prefer lower interest rates.  This reflects the differential mix of net interest payers and recipients.

Chart 4

 

The older person is, on average, just as big a spender as younger persons – on a per person basis (Chart 5).  Spending per person peaks at age 55 to 64.  Thus as the interest income of the over-50’s fluctuates, so will their spending.

Chart 5

 

Conclusions

Monetary policy can be suboptimal if it is made with a poor understanding of the different directions of influence, variable lags, and threshold effects.  Decisions made by the RBA to significantly lift interest rates on the eve of the global financial crisis reflect a poor understanding and their current u-turn on interest rates carries significant risks as discussed below.

Economic commentators can mislead business decision makers and investors if they do not understand the complexities of reactions to interest rates.

The key charts to monitor are interest rate preferences and interest payments as a proportion of household disposable income.  In particular, these identified that a threshold had been breached in early 2008.

Another key factor to monitor is the share market.  The current slump in retail sales growth has been mostly caused by the post-global financial crisis share market recovery faltering in mid-2010.  This caused people who own their dwelling outright, mostly people over 50 who tend to prefer higher interest rates, to dramatically lift their saving rate and cut their spending.  The extra savings have gone into capital-stable investments such as term deposits and cash.

The current round of interest rate cuts may prolong the slump in retail sales growth by further lifting the saving rate of the over-50’s – lower interest rates means lower returns on their savings and consequently a higher rate of saving is required to achieve a given capital target.

Retailers could, if they understood the behaviour of different segments, adapt their segmentation, targeting, and positioning strategy over the course of the interest rate cycle and in response to changes in saving behaviour.

Additional Notes

The RBA and economic commentators were blind to the ills of Australia’s economy through early to mid 2011, despite the obvious signs – see www.charlienelson.com/nelcol/blindseers.htm.

Part of the reason for the RBA cutting rates after talking up rate rises may be to offset the increased cost of bank funding.  The banks may tend to increase interest rates due to higher funding costs resulting from the worsening European sovereign debt crisis.

Update May 2012

A report from the RBA shows that 50% of owner-occupiers with mortgages were ahead of schedule on their payments in 2010.  This figure has varied only slightly over recent years.  Total excess payments were roughly the same as required payments in the December quarter 2011 (Household Balance Sheets, RBA Financial Stability Review, March 2012).

This means that of the 36% of households with a mortgage, only half have their cash flow reduced when interest rates rise.  This is 18% of households, a similar figure to the proportion who want interest rates to rise because they are net recipients of interest (Chart 2).

Update January 2013

The November 2012 retail sales figures have just been released and the commentators continue to display their lack of understanding of how interest rates work:

The failure of the non-mining parts of the economy to respond to the progressive decline in interest rates has also surprised.
David Bassanese, AFR, 7 January 2013.

Shoppers have defied retailers' expectations for increased spending following central bank rate cuts.
Jacob Greber, AFR, 10 January 2013.

Retail figures showed the Reserve Bank of Australia's rate cuts were not gaining traction with households.
Citigroup economist Joshua Williamson, AFR, 10 January 2013.

Lower interest rates have failed to boost the struggling retail sector, as November sales posted a surprise drop on the back of lower consumer spending on non-food items.
Glenda Kwek, The Age, January 10 2013.

If the consumer is cheered by the interest rate cuts, she's keeping it to herself.
BT Financial Group chief economist Chris Caton, The Age, 10 January 2013.

The Reserve Bank's interest rate cut in October had not had a positive impact on consumers' willingness to spend.
AMP Capital senior economist Bob Cunneen, The Age, 10 January 2013.

I wrote to the Reserve Bank about the reaction of consumers to interest rate cuts again in December 2012.  Read my letter.

Home