Fixed Income insights

17 September 2020

The local fixed income market has had a tumultuous year as investors strive to navigate between emergency monetary policy accommodation by the South African Reserve Bank (SARB) and worsening fiscal risks from higher issuance along with collapsing growth adversely impacting on the back end of the yield curve. The SARB has acted swiftly with determination to play its role to cut the repo rate by some 300 basis points since end of 2019. This aimed to help lower borrowing costs to alleviate some of the economic strife resulting from lockdowns imposed by the governments to stem the spread of the Coronavirus (COVID-19).

There has been additional crisis related factors that have also started playing a role in the South African (SA) market that fixed income have had to contend with before. One such factor has been the flow of International Monetary Fund (IMF) funds to SA Government to the tune of $4.2 billion. National Treasury has started drawing down on its sterilisation deposits held at the SARB. These flows, together with hoarding of cash by corporates and other entities, have resulted in a glut of deposits in the SA banking system. This was exacerbated by the lockdown period where economic activity was severely impaired, and spending was put on hold. In part, these developments have conspired with banks having to lower Negotiable Certificates of Deposits (NCD) and short-term funding rates to discourage further cash on their balance sheets which has caused a decline in three month and six month Jibar rates. This has been eroding yield for money market funds posing new and real challenges especially for conservative investors.

As we approach the latter stages of the third quarter we identify four key events and themes/factors that we expect can play a significant role in shaping the path of local yields, the currency and yield curve. These are: Medium-term budget policy statement (MTBPS) [expected 28 October],  S&P/Moody’s SA Sovereign Ratings Update (20 November), United States (US) Presidential Elections (3 November with likely acrimonious and election result delay on mail-in vote count procedures) and lastly vaccine development timeline progress and potential Food and Drug Administration (FDA) vaccine approval near or after US elections in late the fourth quarter.

The sequence of these events is somewhat certain, but exact outcome remains uncertain and hence we opt to take a defensive and cautious stance on local duration given low visibility, expressing an overall neutral duration position. We favour a double overweight in the three- to -seven-year part of the curve (R186 we see trading down from 7.25 to 6.90 area) and keeping exposure in the R2030, R213, R2032 area but prefer a strategic underweighting the bonds beyond R2035 until we have better visibility on local fiscal risks.

The US Presidential election outcome and resultant economic, financial, trade, tax policies are complex to forecast and position for. We are, however, prepared to use any material risk-off or market weakness associated with bear curve steepening to cover underweights. We advocate a flexible approach to make use of cheapening valuations and higher fiscal risk premia embedded in local bonds, which could materialise either before or after the MTBPS, or around the US elections results (which could be contentious and be delayed due to lagged impact of counting all votes cast by mail as all votes have to be signed and all signatures reviewed against the voter roll records). It could also be closer to late November when S&P could pronounce on SA credit rating, where we anticipate a rating downgrade by S&P to occur that will see SA credit rating be reduced by another notch from BB- to B+.

Into next year, National Treasury will be increasingly pressured to return to switch auctions to reduce R2023 outstanding positions and switch holders into bonds further out onto the curve. We anticipate that the prospects and need for additional switch auctions will continue to provide a steepening underpin for South African Government Bonds (SAGBs) curve. Over August the Inflation Linked Bonds (ILBs) have performed impressively well as the market started to recognise a trough in the path of inflation and flip to positive inflation carry in the coming months. Under the assumption of a switch auction programme resumption in the coming months we see R2023 as the source bond yield compress to repo in the 3.25 - 3.5% area from 4.4% currently as this bond is switched by National Treasury.

Its ILB companion the R197 would then still look very attractive at a real yield of 2%. Assuming inflation can average >3%+ in the coming 12 to 18 months, we anticipate that R197 yield can rally to 50 basis points or below. Moreover, should this bond be selected for a switch auction programme to reduce the >ZAR 100 billion redemption exposure in 2023, the R197 yield could collapse to negative territory in a sequel to what happened during 2010/11 period when the R197 predecessor the R189 inflation linked bond was involved in inflation linked bond switch auction programme. We therefore favour ILB exposure in R197 and I2025 space on the curve and stay reluctant to chase the move in long end linkers which have had a stellar run in the past month.

The above nominal bond curve strategy and timing are complicated somewhat by our similarly held constructive view that vaccine development is likely to progress at a rapid pace in the coming months, which together with US elections, can drastically alter risk appetite favourably. United States elections can have an equal if not more impact on increasing or decreasing both risk tolerance and asset and sector rotation and heighten market volatility.

There are currently a handful of vaccines by multiple companies in late-stage phase three trials, the results of which will become known during this month and October. Comments by the FDA, Trump Administration, the healthcare industry and increasing vaccine deals inked between companies and Governments suggest to us that FDA vaccine approval for at least one or more vaccines is likely to occur in the post US election period between US Thanksgiving and early 2021. The timing remains uncertain but once announced we expect it to have strong impact on asset classes and sectors where specifically those underperforming sectors such as hotels, tourism and travel, energy, Office REITS that have been particularly hard hit by the effects of the global pandemic by restricting consumer movement and changing spending patterns and consumer behaviour. We expect that there will be some convergence between sectors such as technology and work-from-home beneficiaries in the months ahead which will be accelerated by the approval of a vaccine (with vaccine rollout to the public likely to start in the US with Operation Warp Speed Vaccine Program launched by Trump administration).

Equally important is that we anticipate that news of a COVID-19 vaccine approval by the FDA will result in a change in portfolio allocation within EMFX from North Asian Currencies (e.g. CNH/CNY, KRW and TWD) into beleaguered currencies where battle to contain the virus infections have had a clear detrimental impact on the economic rebound such as Brazil, South Africa and India. Hence, we expect that a vaccine approval could lead to many emerging market (EM) investors adding BRL, ZAR and/or INR to EMFX exposure on portfolios to benefit from the post vaccine convergence trade. This is expected to be ZAR favourable once we move beyond the upcoming period of fiscal risks, US elections and ratings agency pronouncements. Absence of growth reforms, deteriorating fiscal metrics, ratings agency downgrades and a counter trend rebound in USD into US elections on tightening of the US polls as Trump narrows the race vs. Biden that is in the lead, together with deteriorating fiscal situation in SA result in us having a weaker rand bias. (Chart 1) This view is however offset by a narrowing of the current account deficit and continuing central bank support inciting more risk taking, the high ON FX-implied yields in ZAR FX forwards market that is penalising for EMFX speculators to be long USDZAR which we think help stem the pace of ZAR depreciation.

Chart 1 – Presidential Debates and Trump Election odds will matter for the dollar (First Debate: 29 Sep 2020)

Source: Bloomberg, Ashburton Investments

USDZAR FX ON swaps traded as high as 11.25% implied ZAR rate in August, a whopping 7.75% over repo. In July this traded 2% over and, at the time, we thought it was impressive. Local banks that are long ZAR do not seem interested in lending their ZAR. Perhaps the banks feel the costs outweigh the perceived benefit and they have few short-term USD assets that require funding. While this dynamic persists, it will likely continue to lend support to ZAR. We currently anticipate USDZAR to trade in a range of 15.69 to 17.80 for the remainder of 2020 and like to accumulate USD into bouts of rand strength.

Fiscal risk and debt sustainability – Will MTBPS bring clarity on how SA fiscal consolidation will be achieved?

South Africa faces a fiscal cliff as the total of its civil service wage bill, debt service costs and social grants now exceed Government revenue. A growing number of foreign investors are sceptical whether fiscal consolidation is achievable in SA against the backdrop of the COVID-19 public health and economic crisis, the ongoing legal battle with trade unions, and unresolved financial problems at state-owned enterprises (SOEs). This is evidenced by lack of return of meaningful foreign purchases of local bonds following the liquidation and outflows that occurred in the April to June period when the Coronavirus crisis was at its peak and SAGBs were excluded from the WGBI following Moody’s debt downgrade to sub-IG. According to the latest National Treasury data, foreign ownership of SAGBs and ILBs combined, has declined from 37.3% in January 2020 to 29.9% as at end of August 2020.

There exists scope for foreigners to increase their exposure to SAGBs should SA fundamentals and/or global risk sentiment improve. This is as EM debt funds attract enough inflows due to further signs of a more durable improvement in the global recovery that offers hope for migrating into a post recessionary expansion where EM economies begin to benefit as they emerge from throes of Coronavirus pandemic. It is, however, very important to note that among the contingent of foreign investors not all have capacity to make a meaningful increase in exposure to South Africa local currency debt by purchasing more SAGBs absent large inflows.

The below chart (Chart 2) shows that amongst GBI-EM benchmarked funds monitored by Morgan Stanley, the funds surveyed currently have large overweight exposure to SAGBs which is the second largest long duration contributor in the EM universe after Peru.

Chart 2 – JP Morgan GBI-EM Investor EM Debt Funds – Top Duration OW/UW as measured by Contribution to Duration

Source: Company websites, Morgan Stanley Research

For the upcoming MTBPS the onus will be on Government to illustrate that they are able to deliver on its active strategy communicated during the Special Adjustment Budget presented in June. It presented an active scenario that projects with adequate savings and some revenue measures debt to gross domestic product (GDP) trajectory can stabilise at 87.4% in fiscal year 2023/24.

This active scenario was endorsed and backed by cabinet but entails substantial reduction in government expenses aiming to deliver hard to achieve cutbacks on the civil wage bill. While we have full confidence in Minister Tito Mboweni to present a budget that illustrates the intention to pursue this active strategy that is required to stabilise debt by 2023/24, we remain sceptical whether the market will see the budget as credible. For the active scenario to be delivered upon requires very difficult and severe government cutbacks in expenditures that are politically unpalatable and in medium term threaten social spending programmes. To date it is still unclear if National Treasury has been given a mandate and has full backing and authority to press through with what it takes to deliver on the intended strategy.

South Africa’s seasonally adjusted GDP was released in recent days, plunging by 16.4% in quarter two relative to the previous quarter. This is equivalent to a contraction of 51.0% quarter-on-quarter on an annualised basis, worse than Reuters consensus polling of expectations. Some of the output losses will be partially recovered in the third quarter. However, a national output contraction of this magnitude lacks precedent and will continue to have negative second-round effects on firms and jobs for quarters to come.

When we look at the economic outlook it seems reasonable. Revenue projections, we think it seems fair despite aggressive revenue multipliers, but downside surprises are still possible due to risks to tax collections. There has been a shock to supply side of the economy stemming from the Coronavirus crisis and to the extent that those will be permanent structural factor is unknown. As an outcome coming out of the crisis, we find ourselves dealing with an economic recession, declining potential in GDP, high levels of debt, accelerating interest costs, hardly building assets and below the waterline around the politics of Government - its capacity and efficiency cannot be measured.

From a longer term debt sustainability point of view, it is critical that SA economic growth stagnation is arrested, and a reinvigoration of the economy is achieved. Towards this end many economists, treasury and international agencies have urged SA government to pursue a reform agenda to lift economic growth. To date reforms have been mentioned but very little reforms have been implemented. A recent study by the IRR who does extensive work analysing ANC internal policy documents such as the ANC NEC meeting minutes etc. has been classified in the below chart (Chart 3). The reform agenda analysis has been categorised and currently shows that Governing party policy Counter-reform and No-reforms category currently outweighs the ‘Modest reforms’ and ‘Fundamental reforms’ categories. This analysis highlights the imperative for reform acceleration that is required to change the growth dynamics, which is a necessary condition without which SA debt sustainability cannot be assured.

In his recent Supplementary Budget Speech, Minister Mboweni stated that Cabinet had bought in to National Treasury’s “narrow gate” fiscal consolidation approach. This approach will face several severe political tests in the months ahead, both in the build-up to the mid-term Budget and, more notably, prior to the 2021 Budget Review. National Treasury’s ambition to save ZAR160 billion through containing public sector wage growth faces ongoing resistance from unions. Indeed, the decision to revoke this year’s wage agreement (allowing savings of ZAR37 billion) is being tested in court, with this process delaying the initiation of talks for a new agreement which complicates wage bill certainty over the medium term required to deliver on the narrow gate strategy. Clarity on these critical points of uncertainty will not emerge before the MTBPS.

National Treasury plans to end the temporary COVID-19 additional grant support in October. However, the enormity of the socio-economic consequences of the lockdown will pressure government to retain (or reshape) some of these additional support measures. We view that there is increasing risks for more permanent implementation of these measures beyond October that will result in increased costs for the fiscus, jeopardising the narrow gate strategy. To put into context if these temporary additional grant supports become a permanent feature, it would more than negate the current years targeted savings on wage bill from revoking this year’s wage agreement.

A total of R25 billion has been paid through top-ups to existing social grants, as well as an additional Caregivers Allowance for recipients of the Child Support Grant which would be about R40 billion annual cost to the fiscus if made permanent.

Chart 3 – CRA Analysis of Government and ANC Policy Documents

Source: IRR

When we study the expenditure side it is where things become tricky and difficult. Expenditure this year (fiscal year 2020/21) expressed as %/GDP comes to 37.2%. This is projected to drop to 33.1% in fiscal year 2021/22. This implies a reduction of ZAR220 billion in nominal value which is significant. Once-off temporary COVID-19 expenditures amounted to ZAR35 billion which was financed by about ZAR100 savings identified in the budget. It is unclear how much of this can be made permanent. A fiscal policy expert recently commented that perhaps as much as ZAR50 billion would be feasible. This would then still imply that additional expenditure savings of ZAR135 billion need to be created next year.

While we do not expect Eskom to require additional government funding in the next fiscal year, several other SOEs have reportedly applied for further support (these include the SA Post Office, ACSA, and the SABC). Meanwhile, SAA requires an additional ZAR10 billion to sustain its business rescue process. The political realities outlined in the Supplementary Budget have arguably not been fully calibrated by the ANC and there will be more sustained pressure by some ministers (backed by the social realities that they face in the context of the COVID-19 crisis) to secure additional fiscal resources in 2021 (particularly in light of the scheduled local government elections) in order to accommodate pressing demands.

Does the active scenario stabilise debt? National Treasury said that primary balance will go from about -10% in fiscal year 2020/21 to -2.3% in fiscal year 2022/23. They said that debt doesn’t stabilise in fiscal year 2022/23 but only in fiscal year 2023/24 yet to date National Treasury has not communicated the revenue and expenditure numbers, nor economic forecasts for that year. So as investors we do not know if there is a realistic set of revenue and expenditure projections for that year that is consistent with this debt stabilisation assumption. What is the debt stabilisation primary balance required to achieve this? This is a function of many things, but economists like to refer to it as automatic debt dynamics, which is an interaction between interest rates, growth and the volume of your debt stock. While debt dynamics are complex and non-linear, the evolution of public debt is a function of three things: 1) the existing stock of debt (a result of past policy decisions & implementation); 2) the interplay between growth and debt service costs (influenced by markets and past policy decisions) and 3) the primary balance (directly influenced by policy decisions) all captured in the equation shown below.

Government Debt Accumulation Equation

This equation highlights the complexities and interconnectedness of the dynamics that drive debt accumulation. For instance, for countries with a low stock of debt (dt-1), the primary balance plays a large role in debt dynamics, and govern­ment decisions to run surpluses can generally stabilise debt. For countries with a big existing stock of debt, the rate of growth relative to the cost of servicing debt has a bigger influence, poten­tially increasingly undermined by an unfavourable starting point. Managing this relationship doesn’t guarantee stabilisation without an improvement in the primary balance, but it helps buy time.

The primary balance for fiscal year 2023/24 needed to achieve debt stabilisation for SA remains unknown, but the International Monetary Fund  (IMF) has mentioned in their publication issued with the Rapid Financing Instrument (RFI) loan approval of $4.2 billion that they see the primary balance required of +0.8% (surplus). Other fiscal policy experts have stated that this number may be too conservative and that a primary surplus >+1% may be required. If we use the IMF number as credible that leads us to conclude that between fiscal year 2022/23 and fiscal year 2023/24 there will need for an additional consolidation to the tune of 3% of GDP. Attaching some nominal numbers to that implies another ZAR170 billion of additional consolidation. The scale of this adjustment seems extreme but also undesirable.

Is the active scenario realistic as presented on the narrow path of Matthew 7 cited by Minister Mboweni in his June Supplementary Budget Speech? In our assessment it is not based on the information provided. The statement of intent to achieve the active scenario is an important and a useful one. But the hard work in laying out the practical programme to achieve that intent is still to be done. While we want to consolidate debt and achieve debt sustainability, we also want an economy growing fast enough to provide jobs for the youth. We also need tax revenue to redistribute in society. What is the role of fiscal policy in supporting that? We need a severe and pragmatic spending constraint with focus on the wage bill plus a review of the spending bill with much sharper prioritisation of what matters to us as a nation. We may need to acknowledge that debt is going to increase albeit at a much smaller rate.

In the absence of fiscal consolidation outlined in June, the view is that National Treasury will need to look for alternative financing, especially if foreign investors show no interest in increasing their positions (or worse, continue to sell). Expectations are for regulatory changes forcing local asset owners to buy more government bonds and/or SARB debt monetisation. In the absence of prescribed assets or a change in SARB policy, SA will likely run out of options in the medium term and be forced to a comprehensive IMF package and/or a re-profiling / 'soft' restructuring of SAGBs which involves coupon reduction and maybe maturity extensions. Prescribed assets or debt monetisation in the absence of structural reforms do not solve the problem and only delays the day of reckoning.

Part of the solution must be focusing on the gt (growth) in the equation. Terribly importantly we need a stated and accountable programme of structural economic reform which includes timeframes, details and assignments of responsibility. South Africa finds itself in its 13^th year of load shedding which is a hard constraint on growth and remains unresolved. Fast tracking renewable energy programmes in the energy sector to alleviate electricity supply shortfalls appears to be a logical policy decision. The government will also need to illustrate that it values the role of a dynamic, growing profitable private sector and make policy changes to bring them into the fold to boost investment by fixing the depressed degree of business confidence and policy visibility. South Africa’s fiscal headroom appears to have reached the point of no return where there are no obvious answers how to dig itself out of a dark hole. Revenue levers have been relied upon for many years and has reached a point where the tax base is overburdened. The solution appears to be a combination of some austerity and structural reforms against a backdrop that can find a way to hold steady or ideally reduce debt costs while we see a continuation of quantitative easing (QE) form large DM central banks, injecting liquidity that buys even more time for implementing a solution-based fiscal framework.

“Enter through the narrow gate. For wide is the gate and broad is the road that leads to destruction, & many enter through it. But small is the gate & narrow the road that leads to life, & only a few find it” – TT Mboweni Supplementary Budget Speech, 24 Jun 2020.

US Elections – The biggest single event risk in quarter four  

In recent days ahead of US Labor Day we have seen more evidence of Trump’s re-election chances rising and believe they are higher than currently priced-in. During the month of August, Trump’s re-election odds based on bets bottomed out and have started improving. Currently PredictIt odds have Trump virtually tied with Biden. This is in part due to perceived violent 'pro-Democrat' demonstrations which end up helping Republicans.

Further, we find that polls are likely overstating support for Biden by 5 to 6%. Should US COVID-19 cases decline it may also provide a boost to Trump. Finally, if the debates are cancelled concerns will linger on Biden's possible dementia. While lots can happen to change the odds, we currently believe that momentum in favour of Trump will continue to the extent that by election time the election race will be too close to call, while currently, many investors are positioned for a Biden win. The first of three presidential debates are scheduled to occur on September 29. The second debate is set for 15 October and final debate is set for 22 October.

With roughly 50 days until US elections we intend to further refine our election views. At present in terms of FX, we believe that a Trump 'status quo' (i.e. Trump with a split congress) sees potential for a countertrend rally in USD with as much at 4.5% appreciation from current levels in DXY near 92 to near or above 95. Concurrently, we also think that convergence of COVID-19 equity winners (tech) and losers (airlines, energy, shopping malls, offices, hospitality, etc) is likely ahead of us (i.e. unwind of the “new normal” trade).

A second term for Trump (which is not currently our base case) also increases geopolitical risk premia around Asian assets and increases the odds of randomly-time volatility shocks for all markets. In the table below we provide a summary from Standard Chartered on the two campaigns.

The House of Representatives is expected to remain under Democratic control according to recent research released by Standard Chartered. That said, polling for House races is complex given the large number of districts and factors at play; a Republican upset is possible.

For the Senate, one-third of seats (33) will be contested in November for a six-year term, plus two special elections, for a total of 35. Republicans today hold 53 of the 100 Senate seats. Including the two special elections, Republicans will be defending 23 seats in 2020, against 12 for the Democrats. To flip the Senate, Democrats would need a net gain of three seats if they win the presidency, or four seats if they lose it (the vice president casts the deciding vote in case of a 50-50 split). The latest poll averages show that Democrats could lose one seat and pick up five, for a net gain of four – giving them a 51-to-49 majority control of the Senate.

 (Policy platforms Biden’s “Unite for a better future” vs.  Trump’s “Fighting for you!”)

Source: Media reports, candidates' campaign materials

EM local currency debt outlook and US election outcome scenarios for USD and EM

Real yields within EM appear attractive, but this mostly reflects intensifying credit or macro risks. We believe there is still scope for higher term premium across EM high yielders with signs of rising inflation volatility and fiscal deterioration. With EM local yields rising, we remain cautious on duration in EM high yielders and favour duration exposure in EM low yielders. A few forces are coming together to push EM local debt yields higher. There is a pick-up in volatility in US rates, pockets of inflationary pressure within EM, and diminishing room for monetary policy easing. Moreover, fiscal deterioration is also becoming more apparent in some high yielders like Brazil, India and South Africa. This suggests the onus will now shift more to fiscal policy. In most high-yielding markets there is a lack of domestic appetite for absorbing any additional large expansion in government borrowing. This means that local yield curves could continue to bear steepen. This is also the reason that we remain sceptical of chasing duration in EM high yielders. Granted, yield curves have already steepened in certain markets such as Brazil and South Africa, but we see much more scope for steepening in other high yielders where fiscal deterioration is intensifying, especially India.

Finally, intensifying US-China tensions ahead of the US presidential election also argue for taking shelter in highly rated, low beta markets. We believe that Biden’s lead over Trump is likely to narrow in the leadup to the US elections. Once the results are known (could be delayed for days due to mail-in voting) on the US presidential outcome we also need to consider the Democrat or Republican control of the Senate and the House of Representatives.

There are therefore four possible scenarios of electoral outcomes when considering the US elections and upcoming political establishment into 2021 shown in the below table. These include the Blue Wave, Blue Tide, Thin Red Line, Red Redux outcomes. Each have different implications for USD and EM. Having reviewed the various policies of the two candidates as summarised in the prior section Morgan Stanley analysis reveal their Macro Impact and Sector Fundamental Impact views.

Source: JP Morgan

This suggests that the most positive or bullish outcome for EM would be Blue Tide Outcome. On the opposite side of the spectrum the Red Redux outcome where Republicans have control over House of Representatives, retain control of Senate and retain the White House (while unlikely) currently appears as the most bearish potential scenario for EM in terms of its policy implications.

The final table below provides added high, moderate and low conviction trades based on potential scenarios worth considering. A Blue Wave Scenario is anticipated to be the most bearish for US duration of the various outcomes as it would imply additional fiscal stimulus associated with a huge infrastructure programme, accelerating US growth during a period when the Fed is ultra-accommodative also risk pushing up inflation towards or possibly above the Fed 2% average inflation target.

Table 2: High, moderate and low conviction trades for various US Presidential/Congressional election scenarios



Source: Morgan Stanley