Rodney Dickens thinks respondents are ticking the negative boxes in the ANZ survey to supply the "evidence" to justify OCR cuts

By Rodney Dickens*

At face value the large fall in many components of the ANZ business survey and especially the falls in business confidence and exporters’ expectations provide Governor Wheeler with the evidence he is seeking to justify more OCR cuts, with the potential this will result in a lower exchange rate.

However, it is likely that a number of respondents to the survey are somewhat mischievously ticking negative boxes to increase the odds of OCR cuts and/or a lower exchange rate.

Since 2000 the ANZ survey has been used as a political football by a number of the respondents.

It generally greatly understated near-term economic growth prospects while Labour was in power between 1999 and 2008 and has generally overstated prospects after National took office in 2008. This has greatly undermined the usefulness of the survey as an economic indicator.

I believe it is highly likely some respondents are now ticking the negative boxes in the ANZ survey to supply the “evidence” Governor Wheeler is seeking to justify OCR cuts.

Most of the economists are largely interpreting the fall in the various components of the ANZ survey at face value. Quality analysis would highlight the past “political” bias in the survey and the risk some of the respondents are currently using it to influence OCR decisions and the exchange rate.

If I am right the outlook for economic growth and interest rates will turn out to be materially different from what the ANZ survey suggests and from what the shallow analysis by some of the bank economists suggests.

The ANZ survey has a history of political bias; making it a poor economic indicator

In June the ANZ survey of business confidence recorded a mild, net negative reading of -2.3% (i.e. slightly more respondents ticked the negative box than the positive box)). Based on the ANZ survey, business confidence has deteriorated dramatically since the most recent peak of 70.8% net positive in February 2014 (red line, adjacent chart). Does this mean that in a little over a year the economy has gone from booming to near recession?

The fall in confidence has added to the chorus of calls for OCR cuts and may convince Governor Wheeler to deliver at least one more OCR cut. The economists are largely treating the fall in business confidence at face value. I am highly suspicious the ANZ survey has gone AWOL again, meaning it should be largely ignored as an indicator of the state of the economy.

The chart above shows why the ANZ survey should be considered with considerable scepticism as an economic indicator. From 1993 to 1999 the business confidence survey was a useful barometer of nearterm prospects for annual GDP growth. Over this period the peak correlation was 0.73 with the survey leading GDP growth by two quarters. The red business confidence line has been advanced or shifted into the future by two quarters to reflect the past role as a leading indicator of GDP growth.

However, after Labour came to power in late-1999 and introduced some not so business-friendly legislation in 2000, the survey began to provide a much too negative indication of near-term GDP growth prospects. This is reflected by the massive gap opening up between the survey and GDP growth between 2001 and early-2008. Based on the survey, the economy should have headed into deep recession in 2006 but instead GDP growth improved to 3.5%. Not all business people are National supporters, but during Labour’s terms in office between 1999 and 2008 a significant number of survey respondents were ticking the negative boxes as a form of political protest. Such behaviour is not uncommon with surveys that get lots of media coverage (i.e. the extensive media coverage the survey gets makes it more prone to be used and abused for “political” and other ends).

After National came to power in 2008 and implemented some business-friendly legislation, the survey improved much more than justified by the post-crisis rebound in economic growth in 2009/10 and it consistently provided an overly optimistic indication of near-term economic growth prospects from 2011 to 2014. Some of the respondents were ticking the positives boxes in the survey as a way of expressing political support for National. That is, until Governor Wheeler opened the door to business people ticking the negative boxes in the survey as a means to get lower interest rates and a lower exchange rate.

The ANZ economists correctly pointed out in the commentary accompanying the June survey that the survey of firms’ own activity expectations has been a more accurate leading indicator of GDP growth than the business confidence survey. However, as the adjacent chart shows, while Labour was in power between 1999 and 2008 the own activity survey also generally understated near-term GDP growth prospects, while after National came to power it consistently overstated near-term growth prospects (e.g. based on the survey, annual GDP growth should have surged to around 7% in 2014, but it only reached 3.5%).

Some businesses have clearly answered the ANZ survey with an eye to which party is in power rather than based on what is happening with their business or what they expect to happen in the economy. It shouldn’t therefore be unrealistic to suggest some respondents are now using the survey to try and manipulate OCR decisions. Or, more so, the governor has made it clear he is looking for signs of weaker economic growth to justify OCR cuts that are, potentially more than anything, aimed at getting the NZD down. In the context of the governor’s longstanding campaign to talk the NZD down; it should be no surprise to find exporters providing him with “evidence” that a lower exchange rate is needed.

In June the ANZ survey of exporters’ expectations plunged to be not far above trough levels (black line, adjacent chart). This is despite the tradeweighted value of the NZD against the currencies of our major trading partners (i.e. the TWI) having fallen significantly since the peak levels in mid- 2014. More often than not a fall in the TWI has in the past resulted in higher exporters’ expectations, although the relationship can breakdown when other factors impact (e.g. the advent of the financial crisis in 2008/09 resulted in both falling).

But with leading indicators of international trade for six major countries/blocks having muddled along recently rather than deteriorated significantly (two charts below) and the fall in the NZD TWI improving the international competitiveness of exporters (and local firms competing against imports) significantly, the fall in exporters’ expectations is something of a puzzle. The fall in exporters’ expectations also runs counter to feedback The Herald reported from some export industries about the boost they have received from the fall in the exchange rate (see the link below).

However, it makes perfect sense in the context of the invitation the governor has issued for evidence the fall in dairy product prices will have a significant negative impact on economic growth (i.e. some exporters are ticking the negative box even though the fall in dairy product prices is irrelevant to them because they hope it will help get the NZD down).

I’m not suggesting there is a grand conspiracy, but rather that some business people are acting rationally from a short-term perspective in trying to get down interest costs and the exchange rate.

In the early days of independence there was considerable scepticism at the Reserve Bank about business surveys for these reasons (i.e. the risk the business community would try and bias monetary policy decisions). I know this from my time at the Reserve Bank, including having for a brief period been a member of the Monetary Policy Committee before privatising myself.

My suggestion that the ANZ survey shouldn’t be interpreted at face value is not new. In an April 2011 Raving, after there had been an excessive surge in business confidence that the bank economists largely interpreted at face value, including predicting many OCR hikes, I warned that a significant amount of the upside reflected political bias and should be overlooked.

Given the “political’ behaviour of the ANZ survey since 2010, the first question the economists should ask in response to a significant change, like the fall since February 2014, is whether some of the change was due to factors other than what is happening in the economy (e.g. government policy changes or misguided RB experiments). But that would involve the bank economists being more than glorified journalists.


*Rodney Dickens is the managing director and chief research officer of Strategic Risk Analysis Limited.