A supportive trade narrative can only really be seen on the nominal rather than the real side of the economy – paralleling a currency that remains robust yet an underlying ‘real’ story (of unemployment and inequality) that remains exceptionally weak. Trade gains have been narrowly based, and in an environment of political and regulatory uncertainty these gains are not feeding down to employment or investment growth. Nomura’ analyst Peter Attard Montalto writes
Tuesday 15 August 2017 – Markets continually hunt for a positive story on South Africa and while Q2 GDP will have likely (just) avoided recession thanks to a rebound in manufacturing and mining (off a very low base), there seems to be an excessive focus on trade, which we think is misplaced. In fact, we see real trade underperforming its nominal cousin with the terms of trade (ToT) providing a fig leaf that does not aid the local economy to the degree expected.
Figure 1 shows the consolidation in the headline trade balance since the crisis – in two broad waves from a 2011 surplus in nominal terms and then again since the start of 2016. However, excluding food, oil and gas, aircraft and ships we find that the underlying trade balance is less susceptible to moves in the terms of trade, but has remained stubbornly in a deficit. Indeed, it has not consolidated as much now as it did in 2010 and 2011.
|
|
In Figure 2 looking at the balance by components the key driver of consolidation since 2010 has been in broad mining and minerals, which has been driven only partly by higher metals prices and mainly significantly lower oil prices, especially as ZAR has strengthened. Manufacturing has remained stubbornly sticky after grinding weaker into 2013. Indeed, this is despite the currency losing a third of its value in real terms into end-2015. Meanwhile, ‘other’ goods that reflect consumer products and some intermediate goods have ground slightly weaker. Overall, we can see trade in mining and minerals as the main driver, even as mining volume output has fallen.
Looking in more detail at some of the drivers in Figures 3 and 4 we see that nominal manufacturing export growth (smoothed) has been very weak in manufacturing, despite the solid external demand dynamic. Equally, manufacturing import growth has also been weak since the start of last year as domestic demand and especially domestic private sector investment has steadily contracted. We can see here too the dynamic of the drought with large agricultural import growth through end-2015 and now a mirror contraction, but no real pick-up in agricultural exports, yet despite the strong harvests, suggesting onshore price dampening pressures may remain.
|
|
Broadly, we can see a weak domestic economy in the numbers, but an economy that is failing to take advantage of its external environment because of its domestic-limiting factors. As such, we have consolidated from a trade deficit to balanced trade in recent years almost by accident – ie without a fundamental shift in domestic dynamic or structural reform.
Figure 5 is an interesting look at average mean square errors of our export and import models over different time periods. We can see two interesting things here. First, errors have generally fallen for export forecasts and are now very low in 2017, a wider set of constraints on the domestic manufacturing sector from politics and regulatory issues may well partly be the cause, alongside more integrated global supply chains and perhaps less lumpiness in mining export output. However, imports errors are much larger and have increased substantially this year. Imports have become very erratic, perhaps a symptom of a more volatile ZAR, political uncertainty and a recession. Figure 4 shows that component trends in growth are much less co-synchronous than they are for exports.
Forecast trading on a monthly basis remains highly problematic – our model is based on quarterly real trade, and works backwards to quarterly nominal and then a disaggregated monthly nominal. Errors increase at each stage clearly, but mostly for imports as we can see below.
|
|
Figure 6 shows the stability in the real net trade balance (normalised, both goods and services), it is much slower moving and while not seasonally adjusted here shows much smaller seasonal factors. Nominal trade by contrast is much more volatile here with the volatility driven by large swings in the nominal goods balance (explained mainly by the ToT), but also particularly quarter-to-quarter volatility in services.
The last point here is a Q2 forecast and shows that real net trade remains quite stable, while nominal trade will jump quite significantly higher on a move up in both trade and services. This shows that the market excitement over a large move in nominal outcomes in Q2 will not be reflected on the real side and so the contribution to net growth will remain very low in our view.
Figure 7 shows the contributions of nominal and real trade to nominal and real GDP, respectively. Both were broadly aligned in the pre-crisis period, while since 2010 they have become more completely delinked. Since the start of last year real trade’s contribution has also markedly underperformed that of nominal trade.
|
Fig. 7: Contributions to nominal and real growth of net trade (and 18 month moving averages) |
|
|
|
|
|
Source: SARB, StatsSA, Nomura |
We find that positive monthly nominal trade balances are caused not by underlying shifts from structural reforms or a deeper sustainability in the underlying balances, but instead mainly by shifts in oil prices. We therefore see some support in the balance of payments and partial pressure off of ZAR, but deeper support for growth, job creation and a reduction in inequality is unlikely with these kind of data. It also needs to be considered both next to the more structural and sustainable shifts in low-beta EM trade balances in CEE (to surpluses even stripping out ToT), as well the fact that the income balance in the current account will remain sticky and so the current account is unlikely to be able to be sustainably below -3.5% of GDP in the medium run.