South Africa's 2016 risk report, forecast update - CNBC Africa

South Africa's 2016 risk report, forecast update

Southern Africa

by Nomura 0

Jacob Zuma. Photo: Wikimedia

For South Africa 2016 is set to be a year of economic factors that will drag growth lower still, with rate hikes and raised political risks. In particular, we look for a cornered ANC machine having the possibility of recalling or managing the exit of President Zuma (‘Zumxit’) around July, after a poor showing in the May local elections (though we assign only a slightly lower probability of him continuing until end-2017). However, atop this will be a layer of new, more left-wing policies, including, especially, accelerated implementation of the National Minimum Wage (NMW), though markets will also need to watch nuclear tendering and parastatals. We believe a sub-investment grade rating is possible from one agency by end-2016 for South Africa bonds, while into 2017 there are more likely to be two key agencies assigning junk ratings.

We see no broader shifts in policy that would boost growth during the year, while the SARB attends to increasing inflationary pressures (we revise up our inflation forecast significantly to an average of 6.4% in 2016), rather than just addressing inflationary risks, which was a 2015 theme. We expect only 0.9% growth (vs 1.6% previously) as the economy digests the risk premia impact of the Presidential reshuffle, as well as mining restructuring and the full implications of a global terms-of-trade shock. Following an initial ‘okay’ budget in February against such a weak growth environment, we expect much more consolidation difficulties through most of the year.

The politics

This was not the risk outlook we had originally intended to write! Our usual periodic publication covers events on the horizon and the potential market impact that can often hang only loosely together. 2016 will be a confluence of many different events, but the recent political risk shock with the replacement of Nhlanhla Nene as Finance Minister brings home a central narrative. We think this will be the ANC as a well-oiled political and PR (‘narrative shaping’) machine attempting to battle a very weak economy, loss of support in the local election, weakened but still dominant leadership and a more difficult fiscal position.

The way this machine responds to these shocks and its internal self-stabilising mechanisms will be vital to the outlook for markets and the timing of major political events and shocks. As we advised investors in 2015 around #feesmustfall, the ability for the ANC to take over a ‘bad’ narrative and turn it to its advantage should not be underestimated. The ability of President Zuma to command support, especially in terms of a majority within the ANC leadership, will be equally important in shaping 2016’s outlook.

In 2016 political space in South Africa will be increasingly dominated by more participants, the EFF, in particular, from the start of the year through to the local elections, but so too will unions be important, NUMSA and Zwelinzima Vavi, especially.

To set out the backdrop it is useful to consider the elements of the ANC and the roadmap going forwards.

Figure 1 shows the segments of the ANC that need to be watched – these are the ‘conservative centre-left’ (CCL), who are market- and investor-friendly, transparent and anti-corruption; then there is the ‘left’, which are transparent, anti-corruption, but with policies that are not market- and investor- (i.e., growth and job creation) friendly. Finally there is the mysterious, shadowy ‘tenderpreneurs and cadre-deployee’ segment of the ANC that appear to be non-ideological and focused on rent extraction or power for its own sake.

The balance of these forces is important. First, policy such as the National Minimum Wage (NMW) and foreign land rights ownership, emanate from the left. Policies on SAA and Nuclear come from the tenderpreneur segment; fiscal consolidation comes from the CCL. All three segments likely will be involved in policy in 2016; the issue is how they might influence the other. So on NMW we don’t think the CCL can stop its implementation or reduce the level at which it will be applied, similarly neither the left nor the CCL can stop nuclear tendering moving forwards, a more complex question is whether the left can stop/slow/derail fiscal consolidation as growth slows further (our view is that it can be slowed but not derailed).

All this plays into the power dynamics on succession. In the bubble chart we can see that the two succession candidates – Cyril Ramaphosa and Nkosazana Dlamini-Zuma – currently occupy different parts of the spectrum. However, Ramaphosa’s recent backing by COSATU and SACP suggests he will likely have given away guarantees on policy (NMW, in particular), which may shift him closer to Nkosazana Dlamini-Zuma’s position within the party.

Considering issues such as recall, these power plays and the policy overlay on top are very important for investors to understand.

How The ANC Works – The Bubble Chart

The interplay between the segments, and the fact they are not totally out of whack in size terms with each other, is what we refer to as the ANC’s ‘internal balancing mechanism’. Combining this with the ANC’s ability to use the media effectively could lead one to see the ANC slowly losing power, but maintaining the possibility of a sudden recall or managed departure is possible (‘Zumxit’).The views of the relative size of the different ‘bubbles’ of the ANC has taken a shaking in recent weeks. President Zuma clearly felt he had sufficient banking from the tenderpreneur camp to fire Nhlanhla Nene, and equally that event made us think we had underestimated its size (as occurred in Polokwane when Jacob Zuma was first elected). That said, the flip-flopping and return of Pravin Gordhan has made it clear that the left and CCL could come together and marshal change, and in fact was more imbalance in size terms vs the tenderpreneurs.

That dynamic is important when thinking about political event scenarios for 2016. However, we need to be clear on three different groups of people with different importance (declining order of importance):

ANC NEC and NWC (national executive committee and national working committee), have 100 and 20 people, respectively. We believe that Jacob Zuma has the support of around 60% of the NEC and 70% of the NWC, given how Jacob Zuma came to power in 2007 in Polokwane, and filled key posts around him. It requires more of a shock to shift things for recall, though a managed exit and retirement ‘deal’ may be an easier option. These options ultimately have to come from the NWC and be presented to the NEC, where a majority of 51 members has to be achieved to get recall – again suggesting that a deal may be easier.

ANC elective conference delegates (who will select the next leader at the next elective conference). Within each local ANC branch, up to the provincial level, the interrelationship in the bubble chart above and the balance between factions plays out. COSATU and SACP (who back Cyril Ramaphosa) have no direct representatives except their own members, who are ANC members and who ascend from the local level to be conference delegates. Zuma (Nkosazana Dlamini-Zuma) still have a large sway over the ability to get a majority through this delegate selection process, we believe.

Public electorate – vote ANC or not.

Put simply, we think the NEC, NWC and elective conference will likely retain their positions in the same place within the bubble chart (where Zuma is located), while the public sit more around where we place Mandela, between the two ideological bubbles.

Let us consider some of the key events on the political front:

9 January – ANC’s anniversary speech from President Zuma often contains some policy hints.

January – The EFF will ‘take over’ ABSA offices to make a protest.

Second week January – The first cabinet meeting, most likely the delayed one on the economy that didn’t take place. Some key budget decisions will need to be made.

Last weekend in January – ANC start year Lekgotla, broad attendance from the NWC and NEC and the provincial, caucuses and league leaders. Often gives a read on future policies in SONA.

11 February – State of the Nation Address, varies significantly on its instructiveness, with a lot of policy measures to come in 2016; probably interesting.

24 February – Pravin Gordhan’s first budget speech of his second term as Finance Minister (date not confirmed).

February – NUMSA strike on collective bargaining?

March – Rating agencies all post after budget.

March – Zuma ‘spy tapes’ court case starts?

Q1 – Additional student protest action expected on free education theme (previous was on fee freeze), maybe coincides with the budget or other big events

Q1 – United Front (NUMSA political movement) to have major congress on party formation.

End Q1 into Q2 – Local election campaign on the ground kicks off properly, some (public) polling starts to emerge.

22 May – Local elections pencilled in (date not confirmed).

End July weekend – ANC mid-year Lekgotla.

End October – MTBPS (a Wednesday).

Throughout 2016 – There will be EFF marches, some in Sandton, and some to union buildings in Pretoria or parliament.

June 2017 – ANC policy conference.

December 2017 – ANC elective conference.

What we did not add in the above is the supposed ‘other’ cabinet reshuffle. The focus here is on Tina Joemat-Patterson. As with Nene, noises on this have ebbed and flowed through the year, with no action. While it may be argued that Zuma will be once bitten, twice shy after the Finance Ministry debacle – in a reshuffle he managed to oust Mining Minister Ngoako Ramatlhodi in September, because of the political balancing required from the tenderpreneur faction. His replacement, MJ Zwane, has been problematic for the mining industry, but has not seen a domestic public (or market) reaction at all. Therefore, an additional reshuffle may be possible below the radar.   

A number of things become apparent from the timeline. The first two months of the year present a series of events through which the ANC can define the scope of (supposedly) pro-poor policies like the NMW (and start to implement them through the rest of the year), and then a period of hunkering down and campaigning as we approach the local elections.

We address the specific undercurrents around the local elections below, but if the ANC loses another chunk of voter support as we expect, we then see a period of internal reflection. This occurred after the national elections in May 2014, but this period of refection was too short (lasting until about September, at most) and there was no real follow-through.

The July Lekgotla looks like an interesting time in 2016 for recall/managed exit currents to properly kick in, and as such this should be a very important time for risk in markets. However, we would expect the ANC to close ranks around the president until the point of departure, as happened with Mbeki’s recall. As such, a formalised parliamentary, no-confidence process is unlikely to succeed.

This leaves some issues, however, around succession in H2. The ANC is reluctant to move to an elective conference earlier (we are not sure why – possibly because elective conferences have never been “early” before). H2 may well see some factional infighting between the Dlamini-Zuma and Ramaphosa camps, particularly if Cyril Ramaphosa becomes Interim President. This differs from the period after the Mbeki recall in 2008, when Kgalema Motlanthe became Interim President, in that Ramaphosa is running for leadership and is not liked by half the party – Motlanthe was not a leadership contender at that time and a change in ANC leadership had already occurred at Polokwane at the end of 2007. Motlanthe was Interim President for only nine months; Ramaphosa could end up being Interim President for 18 months (if he loses in the elective conference and is recalled), or almost two years if he lasts until the national elections. As we can see, the process this time is considerably messier.

It is technically possible for parliament to vote anyone else in as Interim President. One possible candidate could be the parliamentary speaker Baleka Mbete (as a compromise between the two succession camps), another is Nkosazana Dlamini-Zuma herself (by an MP standing down and her taking their place in the National Assembly). This kind of optionality illustrates how the current progression of events may differ from the past.

For now, we expect Jacob Zuma to exit around July 2016, Cyril Ramaphosa to take over until the elective conference at the end of 2017, and then Nkosazana Dlamini-Zuma to win with the backing of the dominant tenderpreneur faction and regional alignment among the delegates. Mr Zuma’s exit is far from a high probability event, however, and a close-second scenario is that he remains until the 2017 elective conference, with Nkosazana Dlamini-Zuma winning there. For such a scenario to materialise, there would need to be a lack of shocks big enough to unseat him, the internal balancing mechanism would have to fail to kick in and the ANC would close ranks, thinking it can last to 2017.

We assign a 45% probability to recall/managed exit in Q3 2016 and a 40% probability to Me Zuma remaining until after the elective conference in 2017, with a 15% probability of him continuing right up until 2019.

In conclusion, we think investors should watch for issues around a Zuma exit, who takes over as interim president, and the elective conference, all of which have a policy overlay to drive the political narrative in 2016 for markets.

As we outlined in our last trip notes, we see no call for urgent action to address the threat of a downgrade in terms of wider economic policy to boost growth.

Local elections

There are two factors to watch in the local elections – national share of the vote and key metros. On national share of the vote, the ANC normally does worse than in the national elections and this time we expect a particularly dramatic urban vote squeeze for it, while the rural share of the vote will likely fall much more slowly.

Currently, we model a scenario in which (as opposed to a baseline forecast) the ANC receives around 57.5% of the vote (vs 62% in the last local elections in 2011 and 62.2% in the last national elections) and the DA 29% (vs 23.9% and 22.2%) and EFF each get 8.5% (vs 6.4% at the national elections, they did not exist in 2011).

Such a vote shock may not be enough to remove Jacob Zuma; the loss of key metros would be much more symbolic and embarrassing. In particular, we are watching the DA targets of Tshwane (Pretoria), Nelson Mandela Bay and Johannesburg. Nelson Mandela Bay is the most likely of the major metros to be lost, because the ANC captured only 49.2% in the national election in 2014. In Johannesburg, the ANC received 53.6% last year and 51.0% in Tshwane. It seems unlikely the DA will achieve an outright majority in any of the three (though Nelson Mandela Bay would be most likely, as it got 40.2% in the national elections). We believe the ANC and DA party polling currently shows (before the recent political shock) all three are too close to call, but if the reshuffle shock has the greatest impact in the upper middle classes in the urban areas, these are the sorts of places are where it will be felt, even if the overall share of the vote moves much less.

Fig . 2. Selected Election Statistics

Fig . 2. Selected Election Statistics .2

This leaves the prospect of potentially messy anti-ANC coalitions between the DA, EFF and smaller parties. Talks on this have already begun, but we expect two camps in the DA to lead to indecision. The first camp says whoever the DA forms a coalition with they must show experience running local government outside the Western Cape, the other camp that the EFF is unpredictable. The latter is probably true – the EFF is far better at narrative-shaping than the DA. ANC minority rule may be an option, therefore, or the DA in a vote-by-vote alliance with the EFF. The outcome of this process remains highly uncertain and will depend on what is happening elsewhere.Broadly, we think voter loyalty will remain among the rural and older black urban voters, while we expect dissatisfaction among the younger urban voters, (who demographically form a larger group), which will likely be expressed through not voting than voting for other parties.  


We are fundamentally shifting our ZAR narrative for 2016. Originally (for much of 2015), we commented on SARB hawkishness and hiking offsetting slowly declining credit risk and Fed contagion, envisaging a slight turnaround from the ToT weakness into end-2015.

Our narrative now is of a credit-driven currency moving towards sub-investment grade that stands out as a “shoe dropping” candidate for investors that also shows signs of grinding domestic capital flight and foreigners moving towards underweight. There will likely be more volatility than in 2015, thanks to the Fed and the domestic political story. SARB hikes likely will act as a break to a more disorderly move.

We pencil in 16.0 in USDZAR at end-2016, with upside risks, and 16.0 at end-2017, with further upside risks.

Risks in the very short run are to the downside in USDZAR, thanks to investors seeing value and excess liquidity locally being supportive for the bond curve. The budget will be a key event for the currency and inflation moving higher (core inflation especially through mid-year).

Overall ZAR will still likely appear undervalued on a variety of metrics and that is why the narrative change to credit and risk premia is important. Defining levels too for such a narrative is exceptionally difficult; hence, we think the broad shape of our forecast makes sense, but that the moves and volatility may be outsized.

Figures 3 and 4 show that ZAR around these levels – because of the relative credit stories in Turkey and Brazil – may be a little on the strong side, depending on which RV weighting you want to give it vs Turkey or Brazil. We can see, however, that vs peak Brazil, fear it is not close and is only trading in line with the peak of the Turkey weakness in 2015. With economic fair value metrics of little use (many saying it is significantly undervalued), it makes more sense to track ZAR vs Turkey and Brazil, in particular, or other EM credit stories as they emerge.

South Africa can, in this framework, weaken further through 2016 as the downgrade story is likely to intensify and it moves closer to Brazil in markets’ minds.

Fig . 3. USD Crosses , Rebased To (-1y = ZAR)

Fig . 4. REER’s Rebased To -1y =100

The economy

We are at a difficult juncture in terms of forecasting with some large changes. First, we did not get an opportunity to publish after the recent Q3 SARB quarterly bulletin, owing to the reshuffle, and we need to digest and consider what might be the impact of the reshuffle on growth and the wider economy. We disaggregate these below.

Our growth forecast for 2016 falls from 1.6% to 0.9%. We break down this 0.7pp change, as much is going on below the surface:

-0.3pp is from the breakdown in the quarterly bulletin, in particular the weaker inventory trend, weaker private sector investment and real exports surprising slightly to the downside.

-0.2pp from recent high-frequency data (manufacturing and trade numbers, especially).

-0.4pp is from the impact of the reshuffle onto more negative business sentiment affecting private sector investment, especially.

+0.3pp impact from lower loadshedding assumptions owing to lower demand.

+0.4pp from net trade through the change in ZAR assumptions (muted slightly into REER given higher inflation), mainly driving import compression, though we revise external growth (rest of Africa, especially) down slightly as well.

-0.3pp from more front-loading and then larger overall tightening of monetary policy, with rates reaching 8.00% end-2016.

-0.2pp model specification changes.

We see risks skewed slightly still to the downside, thanks to potential wage round related strike action through mid-2016 that may occur around the manufacturing wage round involving NUMSA, and then the coal and platinum wage rounds starting to kick off into year-end. In addition, the risks of more loadshedding and the downside skew risk in global growth. Our baseline has already taken into account mining sector restructuring and job losses and rates reaching 7.50%. A possible alternate scenario is that of growth around 0.2% if some of these risks materialise in 2016. An upside risk scenario may occur under much more dramatic import substitution and a more buoyant global growth outlook leading to higher metals prices and stronger external demand. Such a scenario might see growth closer to 1.3%, but then the issue would be how much the domestic political issues affects other factors. Figure 6 provides a more detailed view in contribution terms.

We expect growth of 1.8% in 2017, from 1.9% previously, with particular downside risk from strike action.

Fig . 5. Macro Outlook

Fig . 6. Growth Outlook (expenditure GDP)

Our new inflation profile is heavily driven by the shift in ZAR forecast from 13.0 at end-2016 to 16.0. Pass-through assumptions remain for only an extremely gradual recovery in the rate of pass-through over the coming five years from record low (half the long-run average) rates and do not affect the key theme in Figure 7 of a core-breaching target. That is driven by higher retail unit labour cost assumptions and the effects of a weaker ZAR, even with very low pass-through. Food price assumptions have also changed markedly, with maize now over ZAR4,000/MT and still climbing vs previous assumptions of it reaching only ZAR3500/MT. We then expect it to be flat until March 2016 and then fall back to USD3250/MT by end-2017. As has been a continual theme, we lower and flatten the oil price forecast, now forecasting flat at USD35/bl through until June 2016 and then a very slow rise to USD50/bl by the end of 2017 (original forecast was USD45/bl rising to USD70/bl).

The lower oil price assumption has a more marked impact on the peak, which falls in this new forecast from 6.7% to 6.5% (also given some of the recent downside surprises in core). However, the overall impact comes from the weaker ZAR and food prices together, with the underlying assumptions on real disposable income growth remaining steady, a minimal output gap and pass-through. Another major specification of the core equation has also occurred, giving a slightly smoother profile through Q2 2016 that we are happier with now. Core inflation breaches target in July now (May previously). We think specification changes to the food price equations better capture the slow turn we have seen in food and this leads to the second higher ‘wave’ of non-core inflation through Q4 2016 and Q1 2017.

Fig . 7. CPI Inflation Forecast (% Y -o -y)

Figure 7 gives the profile. We now expect average headline inflation of 4.6% in 2015, 6.4% in 2016 and 6.3% in 2017 (5.8% equilibrium view in 2018). This compares with 4.6%, 6.1% and 5.9% previously.

 Fig . 8. Food Price Inflation

Fig . 9. Other Inflation Measures

Risks are skewed to the upside from faster more-aggressive food price pass-through and to the downside in core inflation in the medium to long run, if there is lower real disposable income growth. We think there is an interesting risk to monitor if FX pass-through occurs quicker after a risk premia shock in an environment that should see tighter credit lending.

Notably, this forecast is similar to some of the ‘first alternative’ forecasts the MPC receive from SARB staff that drive inflation ‘fear’ and with it rate hikes. It is interesting to note the above forecast assumes rates rising to 8.00% at end-2016 and then to 8.50%.  

Figures 8 and 9 provide some alternative breakdowns.


Our SARB framework has been remarkably stable since H2 2012 and through the pass over from Gill Marcus to Lesetja Kganyago. However, it will likely ultimately become stressed in 2016, as we see the need to move from neutral rates into tight.

The broad framework will still be intact up to neutral rates of 7.50% – risk management-style hiking, as little as possible, as slowly as possible, in such a very low-growth environment, but the shape of the inflation profile and wage rounds, administered price changes, the skew in risks surrounding ZAR and CPI inflation overall, the drought impact and grinding higher inflation expectations should ensure that despite still weaker growth the SARB can reach a neutral rate of 7.50% by Q3.

We see the first move as a 50bp hike to 6.75% in January to account for the sharp weakening of the currency’s risks into second-round effects, some outflow (though less than might otherwise have been expected from such a political and credit risk premia shock) and with inflation expectations grinding higher.

From there the SARB is likely to continue to put most shocks in the ‘potential’ bucket – i.e., still seeing a minimal output gap. After this if outflows remain moderate and orderly the SARB should be able to revert to 25bp moves until H2 with core and headline inflation breaching target into a wage round period that may cause some alarm for the MPC. As such we see a reversion to 50bp moves to reach 8.00% by November 2016.

Normally, we would expect a pause in an election month like May (the meeting will occur just a few days before), but with inflation risks as they are we think there is a chance of a hike that month; this is a timing factor markets should watch. For now we keep it as a pause month.

We therefore see 50bp in January, 25bp in March, 50bp in July (taking the rate to neutral) and then 50bp in November. We see a final hike (maybe 2x25bp) to 8.50% in H1 2017.

Fig . 10. Real Rates

Fig . 11. The Inflation Fear ‘cycle’

As our forecast envisages inflation with a substantial period outside target for both the headline and core measures, but still returning and remaining within target in 2018, we therefore see only a 100bp into tight territory. We think the degree of fear, however, on the MPC and a view on how much it thinks weak growth is not helping (zero output gap theme) or is helping (demand drag appears) will play a role as well.The move before neutral will require a new framework, but we think the old one will be instructive – sensitive to growth still, but ultimately focusing on ‘inflation fear’ and risk management, looking to cut off second-round effects into core inflation and expectations. Expectation prints and productivity-adjusted real unit labour cost data will be important through H2. Any lack of particular pass-through will cause the MPC to hike much less into tight territory, more disorder (a feared spiral maybe) may cause it to go further to 9.50% say.

We therefore think there is a degree of symmetric risk surrounding our 8.50% terminal rate now. We see cuts occurring through 2018 if inflation stays sticky within target back to 7.50% neutral rate.

The domestic story will dominate the SARB more than the Fed, though we think the MPC will be mindful of the fact the market is pricing is less than the FOMC dots as a reason to skew risks on ZAR and CPI inflation.

Going into tight territory the hawk-dove spectrum will be increasingly important.

Fig . 12. Hawk -dove Spectrum

Fiscal policy

We are not particularly worried about the budget dynamic in the very short run, and the 2016 budget in February may still broadly hang together okay, though will look increasingly frayed at the edges, with large risk holes. Revenue collection momentum remains okay in the current fiscal year, despite the growth slowdown, which we think the National Treasury (NT) will ‘over-assume’ the momentum continuing into 2016/17. A key risk is the credibility of the whole framework, however, if growth forecasts are not revised down significantly. At the time of the MTBPS we highlighted that the NT was already overestimating nominal growth in 2018 by around 1.5pp. It saw 1.7% real growth in 2016 and 2.6% in 2017. We now see 0.9% and 1.8%. This opens a huge gap, with only ZAR2.5bn of contingency reserves in the coming fiscal year and then only ZAR9.0bn the year after. Debt service costs will also need to be significantly revised up on the weaker ZAR and higher T-bill issuance costs as the SARB hikes. Overall, we estimate the total growth and debt service costs-related hole that will need to be filled will amount to some ZAR20bn net of contingency reserve for the budget balance. This can be partly hidden, with inflation measure mismatches between revenue and expenditure (i.e., assuming revenue is stickier thanks to higher CPI inflation, but that government expenditure deflator increases at a slower pace. With NT’s guerrilla warfare with other government departments find cuts, they should be able to make additional savings needed this time at least, while allowing the deficit to expand out a little. Some small tax increases to upper-rate income levels and income band drag and the above-inflation sin tax increases will be required to close that gap. A VAT hike in February seems unlikely before the local elections.  

It will be concerning for markets if the growth estimates are viewed as too high, though strong rhetoric from Pravin Gordhan on the need to be sympathetic to growth and still show commitment to medium-run expenditure may well mute the markets’ reactions. That said, markets may well be a lot more pessimistic, and the ability to keep a primary fiscal surplus in the medium run (as debt service costs rise) will be a very big challenge to watch. The same with debt issuance levels, which will have to glide higher.

But our real concerns focus more on the medium run and the MTBPS at the end of 2016, which may prove more challenging as growth forecasts from the NT are revised down further still, with no contingency reserve and it finds it is much closer to the political bone on cuts. This is why we see key fiscal risk developing more through the year (as markets watch high-frequency revenue data as well) rather than in February.

We think 2016 overall will be a reinforcement of 2015 after the MTBPS – with no space left for countercyclical fiscal policy and a market demanding real consolidation evidence. Ultimately, we fear political room will remain tight, even for Pravin Gordhan, and that cuts to underlying spending will be able to achieve a primary fiscal surplus, but that with wider debt-service costs, the headline deficit will stall to just within -3.6% to -3.8% of GDP in the medium run. This would not be satisfactory for the market or ratings agencies.

However, there seems no scope to ‘double up’ on primary balance consolidation to make space or to offset the lack of growth reforms in ratings agencies’ eyes. Hence, we will see an attempt to use rhetoric to keep the markets onside instead.

Elsewhere in the fiscal theme we have the National Health Insurance funding wrapper white paper from the NT to be published towards end-2016, which we believe will be budget-neutral and propose a VAT hike and employer and employee contributions.

We will also have a detailed contingent liability report sometime in Q3, which will be important for the parastatal risk issue and ratings.

We also need to watch for the DA strengthening its oversight of the budget within the National Assembly, utilising a variety of parliamentary and legal avenues to call attention to the pace of fiscal consolidation as well as the specifics of spending. This may well include further attempts at opposition parties amending the budget.

Other important risk events we are watching


The 2016 Eskom story is likely to be a lot more about finance sustainability than loadshedding in view of our expectations of low growth. Yes, loadshedding will remain a key risk, especially owing to unlucky events (weather, power lines going down, accidents, etc.) and a system that is running very tight with no safety margin.

We await the outcome of the latest RCA (regulatory clearing account) application sometime in January, which will confirm if Eskom’s expected total 2016 tariff increase is 16% (as applied for) or lower (we assume 14%; the SARB assumes 13%). We will then have the substantial MYPD-IV application, which will be for tariffs over 10 years from 2019. The outcome of that process should be around start 2017. At the same time an additional RCA will be made that should be ruled on in H2 2016 and may bring the total 2017 tariff increase to around 15%. Rates markets will likely continue to view tariff events as interesting and relevant for the SARB and monetary policy.

The Medupi and Kusile new build stories will likely remain on the backburner in 2016 – not expected to fully come on-stream until start 2018.


With cabinet approval to move forwards, formal requests for tender should be made early in the new year, with a deadline maybe around early Q2 (pushing back the original deadline of end March). This will proceed, even without the NT signoff on affordability. The announcement of a preferred bidder is expected sometime well into H2. We then expect the start of a very, very long parliamentary and legal obstacle course to begin to try and derail it moving to build.

Markets are still very confused by the issue of nuclear. It is important to remember that it will be, assuming vendor financing, off balance sheet but with the key negative rating implications of massive state-to-state guarantees for repayment of feed-in tariffs – around 25-30% of GDP. Rent extraction through the whole process remains a key concern.


Rain levels remain at exceptionally low levels and the South African Weather Service has warned that Q1 will now be much drier than previously thought, as the El Nino system’s impact becomes more apparent. The longer the drought goes on for the more upside risks to food prices and the greater the fiscal pressure from farmer support. A dry Q1 would be particularly worrying, because it is normally one of the wettest seasons in most provinces and a lack of rain into Q2 would increase the risk of drought and higher food prices will likely last much longer into 2017.

Mining sector restructuring

We watch both the general restructuring plans of the sector (gold, platinum and iron ore especially) and the specifics on the Anglo American asset disposal for the twin themes of major job losses and economic nationalism, and the attempts to create local black-owned mining champions. It is important to realise that these issues are not only (commodity-) cyclical, but are also structural around the lack of competitiveness and cost pressures of the industry.

In line with these wage rounds, starting into year-end, will be particularly interesting; we watch for strike action and AMCU/NUM tensions again, though the bulk of these will probably fall in 2017.

Other wage rounds and strike action potential

We await a major, larger manufacturer wage round after Easter 2016 that will involve NUMSA, but also new COSATU unions attempting to regain lost ground. While strike action is possible, manufacturers seeking to avoid conflict may settle early at high real growth rates. The wage round in 2016 will be unfortunate, occurring at such a time of likely rapidly rising inflation – something the SARB is watching very closely.


We are watching with interest the possible divestment of ABSA by Barclays. We believe that, just as in the mining sector, such opportunities will be used by government to try and push the policy of economic nationalism. This would mean that any private sector buyers would be in competition with a local consortium of PIC (Public sector Investment Company), local development banks and BEE investment funds. This would be a substitute for ‘real’ nationalisation, for which there simply is no money within the state. State-backed retail banking has long been a goal of the ANC, but unachievable within the available resources. This route would allow them to reach the same end goals. Up until now the states’ attention has been focused on the Post Bank, but a mix of poor management and lack of capacity and money has meant such a project hasn’t really gone anywhere in competition with the major banks.

Any outflows by Barclays would be absorbed in large part by the SARB.

National Minimum Wage

We remain particularly worried about the impact an accelerated process of minimum wage implementation would have on the lowest skilled and lowest paid. The political dynamic of implementation around the local elections means legislation with specific levels is likely to come in H1, but jumping the gun vs independent institutions to set such a level dispassionately. A range from ZAR2.1k-4.5k/month seems to have been established by the ongoing debate, but we are concerned it will come at the upper end and so have a more dramatic effect. While outright job losses are less likely, curbs to non-wage benefits and hours may well be the first reaction before lower job creation and more capital-intensive investment is seen. We will be publishing extensively on this topic in 2016.

ABI InBev/SABMiller

The ABI InBev takeover of SABMiller will remain interesting for a number of reasons. First, its relevance to ZAR is limited, even if net inflows are some USD5-6bn. The SARB will absorb this. Nevertheless, the market may well get overexcited. Next the Competition Commission should be watched closely. The Competition Commission will review the deal under its public interest clause and will likely recommend a number of ‘developmental state conditionalities’. These will likely include guarantees on BEE, investment programme targets and job security plans. Finally, the implications for the equity market remain complex, with InBev (old-co) listing in Q1, and yet the deal with SABMiller not closing until towards year-end (after Competition Commission in SA and other jurisdictions).