Lesetja Kganyago: South Africa's growth performance and monetary policy

Address by Mr Lesetja Kganyago, Governor of the South African Reserve Bank, at the Bureau for Economic Research, Cape Town, 22 October 2015.

The views expressed in this speech are those of the speaker and not the view of the BIS.

Thank you for inviting me to address you today.

South Africa's growth rate has been declining persistently since the post-crisis rebound of 2011. The Bank's 2015 forecast has GDP growth at 1,5 per cent, the lowest level since 2009. We expect next year to be only slightly better, at 1,6 per cent. These narrowly positive growth numbers are disappointing, not least because they fall well below the 5,4 per cent growth described in the National Development Plan as the rate necessary for meaningful progress against unemployment and poverty.

I would like to talk to you today about this low growth problem - and about the role of monetary policy.

Boom and bust, recovery and relapse

It is perhaps worth reminding ourselves how we arrived at the present circumstances. The current post-crisis era of low growth was preceded by a large boom. The South African economy expanded continuously from the end of 1998 to the middle of 2008, its output increasing by almost half. Real GDP, on a per capita basis, grew from R44 000 to R54 000.¹ This partly reflected marked structural improvements in the country: a much better fiscal situation, a switch to inflation targeting, and South Africa opening to world trade and financial flows. But, in the mid- to late-2000s, these structural improvements coincided with powerful cyclical forces. Soaring terms of trade and robust borrowing sent growth levels to well above potential, understood as the rate of expansion which does not accelerate inflation. The economy overheated as prices for labour and capital resources were bid up and export prices increased continuously. Moreover, sustained growth in house prices revealed an asset price bubble which had developed in response to low credit costs.

The domestic economy was on an unsustainable trajectory. Rising global food and oil prices - pushed upward as part of the commodity price bubble - added to domestic inflation and prompted greater efforts to slow the pace of fiscal spending growth and a tightening monetary policy stance.² This acted on credit extension and started to slow lending.

It was in these slowing conditions that the Great Recession hit in 2008, following the bankruptcy of the Lehman Brothers. This was an exogenous shock; it was not obvious that it had much to do with South Africa, but it had a major impact on how we have had to reconsider the potential growth rate of the economy. What was our potential growth given the scale of the pre-crisis commodity and house price boom and the post-crisis collapse in global trade? What has been the impact of our electricity constraint? The search for our own 'new normal' has been an increasingly difficult problem for policymakers in recent years and will continue for some time as the economy continues to adjust to the series of shocks that have hit over the past decade.

When the crash hit, South Africa had the leeway to conduct countercyclical policy. National Treasury had fiscal space to keep aggregate demand up as foreign demand evaporated, running significant budget deficits.³ The initial impact on public debt levels was low as a result of the small fiscal surpluses in the last years of the boom.⁴ Monetary policy could also be loosened as inflation was receding rapidly: declining from nearly 9 per cent in 2007 and 2008 to 6 per cent by the end of 2009 and 3,5 per cent by the end of 2010. Vigorous stimulus measures helped output rebound in 2010 and 2011, with growth rates reaching around 3 per cent. Other emerging markets enjoyed similar vigour, and often more. China's growth was in double digits for both years. Brazil was close to 8 per cent in 2010, Russia over 4 per cent. The financial press was filled with speculation of decoupling: the idea that emerging markets could grow by themselves even if developed economies were in crisis.

The vision was attractive but misleading. Emerging markets were actually on the verge of a slowdown. Within a few years, China's growth would be under 7 per cent, and Brazil and Russia would be in recession. From this perspective, 2011 looked like the last year of the boom. But it did not seem that in real time. In South Africa, with potential growth estimated at 4 per cent, the economy seemed to be operating below its normal level even with growth over 3 per cent. The problem got worse from 2012 as growth fell further from this goal. The output gap - the accumulated distance between potential and actual growth - kept growing.

Inflation was similarly uncooperative. We experienced breaches of the inflation target around the end of 2011 and the start of 2012, and again in mid-2013. The case for looking through these breaches was compelling. Each of them could be traced to specific supply shocks, from food or administered prices. The departures from the target were temporary. Core inflation remained within the target range. Inflation expectations were elevated but highly dispersed and reasonably stable. For a flexible-inflation-targeting central bank facing persistently sub-par growth, there was a robust case against responding to these price developments. But that did not make the choice free. The policy had costs. And the benefits disappointed.

One of these costs was shifting to a much higher level of core inflation. When headline inflation first exceeded 6 per cent, the top end of the target range, towards the end of 2011, core inflation was at 3,9 per cent.⁵ From that point, it tracked steadily upwards to a peak of 5,8 per cent in late 2014. It did so in line with repeated, if temporary, deviations of headline inflation from the target range. Headline inflation averaged 5,8 per cent between 2012 and 2014. In these circumstances, it is unsurprising that other, less volatile prices adjusted to that level.⁶

Another cost was the consolidation of inflation expectations around the very top of the target range. As I have already mentioned, average expectations have been fairly stable around the top of the target range for several years now. But focusing on a simple average conceals useful information. Medium-term expectations have converged strongly over the past four years, revealing something near a consensus that South African inflation will be around 6 per cent over the longer term. This is not because of shocks, which cannot be foreseen with any clarity several years out. This is because the 6 per cent is perceived as the normal level.

What of the benefits?

The goal of cutting rates to historically low levels was to close the output gap as inflation was easing. Growth did not accelerate; it slowed persistently: from just over 3 per cent in 2010 and 2011 to slightly over 2 per cent in 2012 and 2013, and a mere 1,5 per cent in 2014. Very low interest rates did not encourage rapid credit uptake in the private sector. Households were overextended from the pre-crisis boom. They benefited from lower borrowing costs, which facilitated debt repayment and permitted some income to be directed to spending. But they did not show much appetite for new debt, except in some relatively small sectors of the overall household credit market.⁷ Corporate finances have been in better shape than households', but businesses also responded unenthusiastically to very low rates. Private sector investment has been weak.⁸ In fact, the strongest movements in credit to corporates have occurred over the past two years, during which we have gradually raised rates.

Meanwhile, the output gap did narrow - but not because growth had picked up. Rather, it was because potential growth had slipped lower. And the inaccuracy of our output gap estimates did not come as a surprise. It is a well-established fact that output gaps are very difficult to estimate, especially in real time. The phenomenon itself is unobserved. There is no single, agreed method for its determination, and competing methods yield varying results. So policymakers do not accept output gap estimates uncritically.

In real time, we thought in late 2013 that the output gap may have widened to as much as ,5 per cent. Since then, we have repeatedly re-estimated the gap. Our latest measures suggest that the gap may have been closer to -0,5 per cent at that time in late 2013.

An intensifying slowdown

This background matters as we contemplate the accumulating evidence of a further growth slowdown, both in the world economy and domestically.

In China, there are new fears of worsening growth. Of course, we have known for some time that China had to slow from double-digit growth rates. We have also known that it was unlikely to need continually expanding quantities of commodities, particularly industrial commodities such as coal, iron and copper. What we have not known is whether this slowdown would be abrupt or gentle. Market assessments have recently shifted, to some extent, away from the benign, gradual scenario. Commodity prices have tumbled as a result; the Bloomberg commodity index fell to levels last seen during the Great Recession in 2008.

There were two triggers for this shift. One was the inflation and subsequent collapse of the Shanghai stock exchange bubble. The second was a devaluation of the renminbi. Exactly what these events mean for China's growth trajectory remains unclear. The IMF's forecasts have not changed from the April to the October edition of the World Economic Outlook: 6,8 per cent for 2015 and 6,3 per cent for 2016. Bloomberg consensus forecasts have finally edged down, in the third quarter of the year, with the median for 2015 now at 6,8 per cent and for 2016 now at 6,5 per cent. That median, however, was constant throughout the period of market turmoil, suggesting that markets remain unsure of the GDP data. It is often argued that forecasts for China's economic growth remain anchored at too high a level. The debate is very active. It is even possible that the data are understating growth by undercounting the services sector and overemphasising the industrial sector, which is plagued by overcapacity. All we really know is that uncertainty over China's prospects spiked in the second half of this year. Along with falling commodity prices, this has fuelled speculation about a possible third chapter to the world economic crisis, with emerging markets taking their turn after the US and the UK, in 2009, and the euro area, in 2011.

These events also affect the advanced economies. The US Federal Reserve declined to raise the Fed Funds Rate recently, despite repeated signals that this course of action was likely. Two Fed governors have since publicly suggested that rates will not rise at all in 2015, although rising rates remains the baseline expectation.⁹ Headline CPI inflation was once again negative in the US in September, although this was once again seen to be temporary. The deflation risk is higher in the euro area, which has faced more outright deflation than the US and continues to have more risks weighted in that direction. Sinking inflation forecasts have prompted a renewed discussion of the possibility of expanding Frankfurt's programme of asset purchases. The fact that the Fed cannot raise rates yet is a sign of global weakness and the European Central Bank must do even more quantitative easing.

Domestic news have also been discouraging. The FNB/BER Consumer Confidence Index is at Great Recession levels. The Barclays PMI has spent much of the year below the neutral level of 50, including through August and September. Our composite business cycle leading indicator has trended lower this year, and is now at levels last reached in late 2009.

Unexpectedly, the economy contracted in the second quarter of the year, shrinking by 1,3 per cent, with negative contributions from agriculture, mining and manufacturing. But this number should not be treated as cast in stone. GDP estimates are necessarily uncertain. We are often reminded of this by the US, where the practice is to release preliminary estimates of growth followed by second and final versions - and then sometimes a revised estimate after that. In the first quarter of this year, for instance, US growth was described as 0,2 per cent, then -0,7 per cent, then -0,2 per cent and finally - I hope - 0,6 per cent.

Similarly, calculations of South African output are not straightforward. The second quarter of this year is a particularly striking example. The estimate of South Africa's GDP growth taken from the demand side gave a different verdict to the supply-side version. The demand-side number was positive. The gap or residual required to reconcile the demand- and supply-side versions was 2,2 per cent of GDP. This is very large and suggests that revisions to the second-quarter number will be significant.

Our forecasts, like those of others in the market, showed weak but positive outcomes for that quarter, and it is possible that these will turn out to be somewhat less wrong than seen at first sight. We do anticipate growth to be positive in the third quarter, but this is scarcely comforting. Growth will remain very low over the next few years. Our best judgment is that potential growth for 2015 has fallen all the way to 1,8 per cent. And with actual growth expected to be 1,5 per cent, the output gap will increase somewhat.

Policy and the problem of stagflation

Let me turn now to our current policy stance. In this environment of weak economic growth, we are sometimes encouraged, as the monetary policy authorities, to try and do more for growth. This is a valid and important contribution to policymaking, so I would like to spend a few minutes explaining the monetary settings we have in place now and why we think these settings are supportive of growth.

The inflation outlook is not favourable. We expect headline inflation to average more than 6 per cent in the first and fourth quarters of next year, and just less than 6 per cent in 2017. This forecast faces sizeable risks, especially from currency depreciation as well as wage and price determination processes. With weak commodity prices and US monetary policy normalisation coming closer, we cannot be complacent about the exchange rate and its potential inflation consequences. Furthermore, we confront medium-term inflation expectations bunched around the top of the target range. The risk of positive inflation shocks feeding into higher expectations and price setting remains very high.

Were the forecast more favourable or were inflation expectations anchored firmly within the target range, monetary policy would enjoy greater freedom of action. If wage and price setting in the economy were more attuned to weak demand and the usefulness of lower inflation, we would have more policy space. But we do not have that space. It has been eroded trying to get growth back up to potential. Now we have to contend with significant risks of a sustained breach of the inflation target range. For this reason, we entered a tightening cycle in January 2014, which has so far brought the repo rate 100 basis points higher, to 6 per cent. The real rate has been slightly negative and is now about zero, relative to an historical real rate averaging around 3 per cent. This gradual and limited rise, occurring over nearly two years, has shown due concern for growth while maintaining our attention on a rising inflation rate.

Some voices say that this is a mistake: that we should let inflation go and focus on the output gap. This advice is problematic on several levels. As economic literature has repeatedly demonstrated and as our experience has once again demonstrated, the output gap is an unreliable measure.¹⁰ It is a poor guide for policy at the best of times. It is an even worse reason to depart from the guidelines entirely. It is possible that cutting rates would boost short-term growth, but we saw little evidence of that before the commencement of the tightening cycle. We should also note that with potential growth so low and structural constraints binding, additional monetary accommodation is likely to generate more inflation. Rigidities in product and labour markets mean insiders would push wage and price demands higher, limiting the gains we think could be achieved with looser policy. The stickiness of expectations has been particularly obvious lately, with oil prices suppressing headline inflation all year but medium-term expectations apparently unmoved. Evidently, inflation expectations are not only backward-looking, suggesting that expectations are already stickier than they should be. A more sustained and general rise in inflation expectations would be difficult and expensive to reverse.

By anchoring expectations within a credible inflation-targeting regime, we prevent inflation creeping higher. This is a valuable objective. Additional inflation would hurt people who lack the power and privilege to protect the real purchasing power of their wages and savings. Higher inflation is also a medium-term threat to competitiveness. World inflation is currently unusually low. A number of major economies are close to deflation. This expands our inflation differential with our trading partners and competitors. The floating currency offsets the problem but not in a reliable or permanent way. Instead, whenever nominal exchange rates stabilise, the inflation differential helps the real exchange rate to appreciate.

To defend the inflation-targeting framework and emphasise the costliness of inflation is not to ignore growth. Indeed, monetary policy in South Africa remains extremely accommodative. Interest rates have gone up but they are not high, nor are they climbing rapidly. The repo rate is still close to zero in real terms. This is a low level from a historical perspective. It is also low compared to peer countries. Monetary policy can be so supportive of growth precisely because of the flexible-inflation-targeting regime. We have the leeway to look through shocks and focus on second-round effects. For instance, we do not need to try and control the exchange rate, a difficult task at which we have traditionally failed. Rather, we can respond to the potential for depreciation to feed into inflation, and specifically wages and inflation expectations. This is why, for all the shocks buffeting the economy, not least a large depreciation of the rand, we have adjusted the repo rate just a quarter of a percentage point this year.

Conclusion

South Africa's growth has slowed steadily since 2011, despite expansionary fiscal and monetary settings. These policies were designed to restore growth to its potential. Unfortunately, our potential growth rate has declined. We anticipate that growth will start to improve over a two- to three-year horizon, as potential rebounds. One important change will be a relaxing electricity constraint as new sources of supply feed into the grid. Improving world growth will also tend to support the domestic economy, and it is in this context that we need to be sure that inflation or other obstacles do not deprive us of market access or competitiveness.

Over the next few years, monetary policy will contend with high inflation. We are not in the position of the major economies in Europe, North America and Asia, which are having trouble lifting growth and also in generating inflation. Our problems are more prosaic, and are unlikely to require major revisions to the textbooks. In keeping with the flexible-inflation-targeting framework, monetary policy will be set to keep inflation within the target range over the medium term. Fortunately, the credibility of our regime will help us to do this in a way that is also supportive of growth.

Thank you.

---