A winter of discontent looms as South Africa remains firmly in the throes of its harshest period of loadshedding yet.
Our besieged economy has only been spared from loadshedding for one day this year, with a cumulative 13 000-gigawatt hours (GWh) of electricity supply lost in the year-to-date period – already more than the cumulative electricity supply lost to loadshedding in 2022. This equates to an average Stage 4 loadshedding across the country for the period, with no relief on the horizon as the winter chill of the southern hemisphere sets in.
In recent years, the colder months of the year (June – August) have seen an average 3GWh increase in electricity demand, largely attributable to the use of electrical heating by South African households and businesses. While previously buffered by a decrease in planned maintenance across Eskom’s electricity generation fleet, the current unfavourable tilt between planned and unplanned maintenance across its hobbling generation fleet suggests that we should only expect limited relief from the cyclical supply and demand dynamics this year.
Consumers pay the price
Figure 1: Energy Availability Factor (EAF) & planned/unplanned maintenance
Source: Eskom, Futuregrowth
This backdrop presents two primary risks to the domestic bond market, the first via monetary policy and the second via fiscal policy. From a monetary policy perspective, the costs associated with running backup power are significantly greater than baseload generation, contributing to upside inflation risk. The Council for Industrial Research (CSIR) and the South African Reserve Bank (SARB) estimate the cost of running diesel generators (R6.75/kWh for 2022 assuming an average diesel price of R22.51 per litre) to be 133% greater than the weighted average cost as sourced from municipalities across the country (R2.90/kWh for 2022). The current leaning by industry towards diesel generators to mitigate loadshedding risk, the costliest of the electricity back-up solutions, combined with a backdrop of strained corporate profitability suggests that the additional costs associated with running a business in this environment are being passed on to consumers. This much is evident in the elevation and stickiness of consumer price inflation (CPI) in the past year. The SARB indicates that the cost-push pressures associated with loadshedding raised the average CPI rate by 0.6% year-on-year in 2022, and estimates an additional 0.5% year-on-year uplift in 2023.
The recent trading statement by Pick n Pay Limited highlights the extraordinary costs incurred by corporate South Africa to keep the lights on in the second half of 2022, when loadshedding intensity ratcheted significantly higher. This company alone spent R346 million on diesel charges over the period. While acknowledging that the energy needs of corporate South Africa vary significantly across companies and industries, accounting for the modest 0.10% that Pick n Pay contributes to the JSE All Share Index (ALSI) nonetheless begins to highlight the significant costs borne by business and the economy due to Eskom’s inability to generate sufficient electricity supply to meet current demand, let alone meaningfully grow the economy.
A perfect storm
From a fiscal perspective, weak macroeconomic growth undermines income tax revenue. Of the three main income tax revenue sources, personal income tax (PIT) is the least impacted by variable macroeconomic growth outcomes. This is explained by the concentration and relative stability of South Africa’s personal income tax base. Value-added tax (VAT) and corporate income tax (CIT) receipts are positively correlated to macroeconomic growth outcomes. This positive correlation, strongest at an eight to nine-month lag, has a net negative impact on aggregate income tax revenue in a low-growth environment.
While fiscal revenue outcomes have historically exhibited a strong and persistent procyclicality with real Gross Domestic Product (GDP) outcomes, fiscal expenditure on the other hand has historically had a convex relationship to real GDP outcome – exhibiting a strong positive relation to real GDP growth both in upswings and downswings. What this results in is fiscal expenditure pressure during both good and bad times, and stability when real GDP outcomes are closer to the historic 2% year-on-year trend rate. Moreover, although the GDP deflator has historically contributed to inflating fiscal expenditure outcomes, it has lent very little support to fiscal revenue receipts. Taking this sum of the parts approach, what we have then from a nominal GDP perspective (real GDP plus the GDP deflator), and at the very heart of our concerns around South Africa’s fiscal sustainability in this stagflationary environment (low real GDP and high inflation), is a near perfect storm for the fiscus to weather.
South Africa’s growth challenge
Figure 2: Revenue/expenditure growth relative to real GDP growth
Source: National Treasury, Futuregrowth
Already anaemic macroeconomic growth will be further strained by the electricity deficit. Some of this downside risk will be negated by the significant investment by local business in renewable back-up electricity generation which has and will continue to contribute to the improved energy efficiency of the economy. Despite this, and consistent with the SARB’s estimates, our modelling suggests economic growth downside of -2.2% for the calendar year, assuming a constant electricity supply deficit of 4GWh (equivalent to Stage 4 loadshedding). The risks to this assumption are twofold: the first is the clear potential for higher stages of loadshedding as the winter cold begins to bite; and the second is the risk of exponential effects in the relationship between the electricity supply deficit and lost output.
Turning to economic growth, South Africa’s growth frailties are well documented, with a perennial inability to consistently grow beyond 2% in real terms. It is, however, when measured relative to peers from a GDP per capita perspective, that the extent of the growth challenge becomes clear. Measured on this basis, South Africa is a relative underperformer regardless of the chosen peer group.
Figure 3: GDP per capita
Source: World Bank, Futuregrowth
This backdrop, combined with elevated debt-servicing costs and growing current expenditure demands crowding the space for growth-enhancing fiscal spend, serves as a clear binding constraint on economic growth. The tax revenue windfall from the recent COVID-induced commodity super cycle provided much-needed breathing room to the fiscus, contributing to the first main budget primary surplus in 14 years - but this benefit is now firmly behind us. What is increasingly apparent is that economic growth, preceded by difficult but necessary policy choices, is the only way to heal the country’s fiscal position. Without growth, and given the current fiscal inertia, a debt trap is inevitable.
Impact on the bond market
Figure 4: Real GDP vs Debt/GDP
Source: Futuregrowth
This backdrop, combined with elevated debt-servicing costs and growing current expenditure demands crowding the space for growth-enhancing fiscal spend, serves as a clear binding constraint on economic growth. The tax revenue windfall from the recent COVID-induced commodity super cycle provided much-needed breathing room to the fiscus, contributing to the first main budget primary surplus in 14 years - but this benefit is now firmly behind us. What is increasingly apparent is that economic growth, preceded by difficult but necessary policy choices, is the only way to heal the country’s fiscal position. Without growth, and given the current fiscal inertia, a debt trap is inevitable.
Market pricing has adjusted in recent months to better reflect these risk dynamics, with upward pressure on both the level and slope of nominal bond yields. The effects of a steep monetary policy tightening cycle from the SARB and its persistent hawkish bias will contribute to tempering price pressures, but we don’t expect a meaningful reduction in the sovereign risk premium embedded in bonds until macroeconomic growth prospects meaningfully improve. We express this view in the domestic bond market with our overweight allocation to 12–20-year maturity nominal bonds, capturing the accrual benefit but limiting the interest rate sensitivity associated with longer-dated bonds.
