Implications of SARB targeting 3% inflation
Conjecture about the lowering of South Africa’s inflation target from its current 4.5% has gained momentum since the publication in February 2024 of National Treasury’s Macroeconomic Policy Review[1]. The review argues that inflation targeting’s goals have “broadly been achieved…but some adjustments to the framework may be desirable given inflation differentials compared to our peers and trading partners”. This has become more topical since the SARB governor mentioned in April that policymakers were in discussions regarding lowering the central bank’s inflation target before we get to 2025.
Given that the target is 4.5%, South Africa’s inflation rate remains above the global average (4.0%) and its main trading partners i.e. Germany (2.0%), US (2.0%), China (2.5%) and UK (2.6%), except for India (7%). In accordance with the principle of purchasing power parity (PPP), this persistent differential requires the steady depreciation of the rand’s exchange rate, creating a feedback loop as a weaker currency puts pressure on inflation.
Since the introduction of an inflation targeting monetary policy framework by the South African Reserve Bank (SARB) in February 2000, inflation has trended around a lower mean with reduced volatility. Research produced by the SARB also shows that exogenous inflation shocks (for example a higher oil price) have been contained more effectively since 2000 such that “inflation has reverted to target faster and at lower interest rates. This is because inflation expectations have been better anchored.”
Figure 1: Decomposition of SA’s bond yield
Source: Bloomberg, Matrix
Fiscal Implications
By protecting the value of the currency, inflation targeting has significant fiscal advantages. Theoretically, sovereign bond yields need to compensate investors for a sovereign’s default risk over and above the global risk-free rate. Local currency bonds must provide additional yield for currency or inflation risk (Figure 1). The review argues that South Africa’s fiscal policy has become structurally unsustainable at current and forecasted growth rates. This has increased the credit/default risk priced into SA bonds. Thus, since credit risk is unlikely to structurally compress in the medium term, it is up to monetary policy to do the heavy lifting to counter macroeconomic risks. Lower, less volatile, and more predictable inflation compresses the currency risk premium embedded in local currency sovereign bond yields. It should also reduce the term risk premium by reducing interest rate uncertainty. A lower inflation target can therefore lead to lower long-term borrowing costs and reduce the cost of servicing inflation-linked bonds. Lower more predictable inflation and lower borrowing costs could also encourage investment. Treasury’s inflation-linked expenditure would also benefit; we estimate that inflation of 3% versus 4.5% would reduce the wage bill by R60bn over the current Medium Term Expenditure Framework (MTEF).
What is the correct rate of inflation?
Surprisingly, apart from matching trading partner rates of inflation, no economic research has determined the optimal inflation rate. Dusting off the Cobb-Douglas production function, it predicates that over the longer-term inflation is a function of population growth and potential real GDP – as measured by labour and capital utilisation rates and total factor productivity (TFP)[2]. TFP broadly refers to levels of technology and productive efficiency (Figure 2). The relationship between inflation and population growth is positive and between inflation and TFP is negative. For example, the introduction of the internet raised TFP and lowered inflation because output increased for any given quantity of labour and capital. SA’s population growth is estimated at around 1.6% and potential real GDP growth is estimated to have declined to below 1.0%. Fundamentally growth in demand outstrips growth in productive capacity, which has been in decline.
Figure 2: An increase in Total Factor Productivity (TFP) can increase growth while reducing inflation
Source: Matrix, AFDB
National Treasury’s review advised that more technical work on an appropriate level of inflation for South Africa should continue. But, if trading partner equivalents are a guide, then the target should be lowered to 3.0%/3.5%.
Is it practical for SA to lower its inflation target when headline CPI has been above 4.5% for almost 80% of the time (excluding 2020) since it was explicitly targeted in 2017? Over the past 5 years, headline CPI has averaged 5.0% and administrative price inflation has averaged 7.2%, such that if we exclude administrative price inflation CPI has averaged 4.5% – exactly on target. It is likely that South Africa can achieve a lower target excluding administrative prices, which by their nature are not responsive to monetary policy and require political buy-in. From a theoretical perspective, political buy-in would allow South Africa to increase its TFP.
Impact on interest rates
Lowering an inflation target needs to be a gradual process. Using the SARB’s current Quarterly Projection Model (QPM) Taylor rule parameters and forecasts, estimates that the policy rate can be cut by 44bp by the end of 2024 and 80bp by the end of 2025. Lowering the inflation target to 3% gradually over a period of 10 quarters would also allow for cuts, but these would be far more gradual, at first. The gain associated with this initial pain is that the terminal rate is lower; instead of reaching a steady state at a 7.0% neutral policy rate (2.5% real neutral rate plus 4.5% target) it would be reached at a 5.5% neutral policy rate (2.5% plus 3.0% target).
We provide a stylised model using the SARB’s QPM Taylor rule in Figure 3. The policy rate remains above its 4.5% inflation target profile until mid-2027, after which it falls rapidly as inflation is assumed to respond, with a 12- to 24-month lag, to higher rates for longer, and move towards a 3% target. In the model we assume that inflation does not actually reach 3% over the forecast period (to mid-2028) but does trend below 4.0%. We note that, in reality the target is also flexible, which means that temporary deviations from the target are acceptable provided that inflation returns to the target range over a reasonable period (usually one or two years).
Figure 3: Repo rate simulation using SARB’s QPM Taylor rule under different inflation targets
Source: SARB, Matrix
The SARB is working on its next 5-year Strategy Plan, to be launched early next year and we assume that it would include a lower inflation target. The announcement is formally the responsibility of the finance minister. It will likely be announced when he appears in Parliament and may be at the presentation of the Medium-Term Budget Policy Statement (MTBPS) later this year or the February 2025 budget, which is probably more likely.
Interestingly, when the US Federal Reserve (Fed) implemented a 2% inflation targeting framework it decided not to make the target public: “Greenspan simply did not want his discretion constrained in any way when it came to possible actions, he might want to take…there was “an unwillingness to create an output gap to get to 2 percent” during the periods when inflation was above that. “If you make 2 percent public, and you’re running at 2.5 percent, then the question is, ‘why aren’t you creating unemployment to get to 2 percent?’ That’s not a position anyone really wanted to be in.”
The decision to lower the target needs to be reached by consensus between representatives of the SARB and the National Treasury. We expect this would be a process of robust debate and another opportunity for the SARB to display its independence.
[1] https://www.treasury.gov.za/documents/national%20budget/2024/Macroeconomic%20Policy%20Review.pdf
[2]https://www.afdb.org/fileadmin/uploads/afdb/Documents/Publications/WPS_No_263_Trends_in_Factor_Productivity_Efficiency_and_Potential_Output_Growth_in_South_Africa_Za.pdf
