South Africa: FX and rate call change
This post is a guest contribution by Michael Kafe and Andrea Masia of Morgan Stanley.
In last week’s FX Pulse, we revised our USD/ZAR outlook for end-2011 to 7.0 (7.50 previously) and end-2012 to 7.50 (8.0 previously). The improved USD/ZAR outlook should combine with the recent decline in oil prices to place a soft lid on domestic inflation. We now expect CPI to hit 6.0%Y in December 2011 (previously by September 2011), and peak at 6.1% in February 2012 (previously 6.3%Y) before coming back to close 2012 at 5.7%Y (previously 6.0%Y). For 2013, we look for an average CPI print of 5.7%.
Also, we believe that the MPC is likely to extend its inflation forecast horizon to 2013 at the upcoming July MPC, where inflation is likely to return more meaningfully into the target band. Such a return to a ‘hump-shaped’ profile should allow the MPC to comfortably revert to an argument that the expected breach of target is mostly driven by exogenous factors that eventually fall out of the wash, thereby obviating the need for tighter money.
Against the above background, we are now inclined to believe that the SARB is likely to delay policy normalisation until at least 1Q12, where we look for the first of four successive 50bp rate hikes, taking terminal rates to 7.5% by September 2012. However, we must point out that, while the SARB may prefer to engineer successive, measured rate hikes, any soft patch in growth – be it local or global – could force it to pause along the way.
Revisiting Our USD/ZAR Forecast
Together with our FX strategy colleagues, we took the opportunity in last week’s FX Pulse to revisit our USD/ZAR forecast (see FX Pulse: Offering Sterling, July 8, 2011). Our call at the beginning of the year was for USD/ZAR to remain within a trading range of 6.80-7.20 in 1H11 before weakening to 7.50 by year-end, thanks to our view of a widening current account deficit, a narrowing of interest rate differentials, rising domestic inflation and a likely slowdown in capital inflows. For 1H11, the current account deficit was somewhat smaller than we had expected, while capital inflows were significantly higher than we had hoped for. It is therefore not entirely surprising that USD/ZAR closed 1H at 6.75 – stronger than our forecast of 7.00.
Looking forward, although we maintain a weakening bias into year-end, we have turned somewhat more constructive than before. For the coming few months, we expect the currency to be resilient, but with our global FX team expecting EUR/USD to fall to 1.36 by the end of the year, we see scope for a move higher in USD/ZAR to 7.0 by December 2011. As for 2012, we continue to expect the ZAR to underperform the rest of EM, EUR and USD. We take this opportunity to elaborate on our change in view and its implications:
A Healthier Balance-of-Payments Position
Our revised estimates for the capital account of the BoP show a more constructive funding environment than we had believed would be the case earlier in the year. For example, we now see the basic balance on the BoP registering R21 billion in 2011 (R12 billion previously), and -R23 billion in 2012 (-R35 billion previously). This should allow for healthy FX accumulation of as much as R42 billion in 2011. Here’s why:
First is the current account deficit. Despite the strong rebound in both household consumption expenditures and gross domestic capital formation, South Africa’s 1Q11 trade deficit came in at R22 billion – much higher than our forecast of a R4.5 billion deficit – thanks to the combination of buoyant commodity export inflows and weaker-than-expected import growth. Looking forward, we believe that continued buoyancy in commodity prices is likely to place a floor under export proceeds, while a potential deceleration in consumer spend (mostly technical) helps to place a lid on import growth. We have therefore revised our full-year current account deficit marginally from 4% of GDP to 3.7% this year, and from 4.9% to 4.5% in 2012. The lower deficit path should help take some pressure off the currency, in our opinion
Second is the capital account. For 1Q11, although portfolio equity inflows were dismal, fixed income inflows turned out to be much stronger than we had hoped for – thanks to the global tide of fixed income interest in emerging markets generally. Unrecorded transactions also posted a solid R29 billion print that was well ahead of our -R5 billion expectation. For the quarter as a whole, net capital inflows were north of R30 billion – much more than our R7 billion forecast in January (see South African Chartbook: Steady Growth with Higher Inflation, January 31, 2011).
Looking forward, we believe that fixed income inflows are likely to continue to trickle in, particularly once the market starts to see through the inflation peak for real. With regards to equity flows, we look for an improvement in 2H11 compared to earlier estimates, as our equity strategists have turned more constructive on emerging markets, and are now forecasting 15% upside in MSCI EM for the remainder of this year (previously 11%). Also, contrary to our expectations of a dry-up in FDI inflows in 2H11, Chinese mining group Jinchuan made a R9.1 billion bid for South African miner Metorex on July 5, 2011. This deal alone – if successful – equates to just under half of the 1Q11 current account deficit. Other flows such as banking sector FX repatriations, which played a significant role in funding the 1Q11 deficit, could come in too. On the whole, therefore, it appears that South Africa may find it less difficult to fund its current account deficit than we had previously thought.
Third are interest rate differentials. At the time of our previous ZAR forecast run in January, our US economists were looking for policy normalisation as early as 3Q11. Our view at the time was that South Africa would lose its relative yield appeal if that were to occur. Our US economists have since changed their call, and have pushed out their expectation of US policy tightening to 3Q12, allowing ZAR yields to maintain their relative yield attractiveness for longer than initially expected.
With spot USD/ZAR at 6.90, our fair value model suggests that the ZAR is some 17% overvalued relative to its 3Q11 estimate of 8.15. We must highlight, however, that the balance-of-payments analytics discussed above suggest that the currency may remain overvalued for the time being, thanks to the huge tide of capital inflows.
Thus, although we maintain a weakening bias on the currency, we believe that the extent of such weakness will cap USD/ZAR at 7.00 by end-2011, and 7.50 by end-2012.
Downward Revision to Inflation Forecasts
As we have demonstrated in previous research (see South Africa: Another Look at Oil’s Influence on CPI, March 14, 2011), every 10% fall in oil prices shaves roughly 0.2% off South Africa’s CPI. So far, the recent move in Brent oil prices from US$120/bbl to US$106/bbl in June led to a 33c/l drop in the July domestic regulated price of fuel. This, together with the improved ZAR outlook, has led to a downward revision to our inflation trajectory. On the whole, we now expect CPI to reach 6.0%Y in December 2011 (previously by September), peak at 6.1% in February 2012 (previously 6.3%Y) and decline to 5.7%Y by the close of 2012 (previously 6.0%Y). For 2013, we look for an average CPI print of 5.7%, although inflation should fall towards 5.5%Y by mid-year before rising again in 2H13.
Core CPI Forecasts Broadly Unchanged However
With regards to core CPI, we expect CPI excluding food and energy (CPIXFE) to rise from current estimates of 3.8%Y to 4.0%Y in June, and to cross over the mid-point of the inflation target band to print at 4.6%Y in August, before peaking around 5.7%Y (previously 5.8%Y) in 3Q12. Thereafter, we expect CPI excluding food and energy to flat-line at about 5.7%Y (i.e., just below the upper end of the inflation target band) into 2013.
But close to half a percentage point of the increase in CPIXFE is simply due to electricity tariff increases – an important exogenous factor that monetary policy is unable to influence directly. On our estimates, CPI excluding food, energy and electricity (CPIXFEEL) is likely to rise from 3.3%Y in May to the mid-point of the target band in December 2011 (previously September), before leveling out at around 5.2%Y in 2H11. For 2012, we expect CPIXFEEL to average some 5.1%Y. This stricter definition of core CPI portends a less-worrisome outcome for second-round inflation.
Finally, CPI excluding administered costs (CPIXAD) is likely to rise from 3.5% in May to 3.9% in June and is likely to cross over the mid-point of the target band in September. Thereafter, we expect CPIXAD to peak at around 5.5% in 1H12, before falling back to 5.3%Y by the close of that year. CPIXAD should converge with CPIXFEEL at some 5.1% over 2013, we believe.
Implications for Policy
At the last MPC meeting on May 12, 2011, the SARB’s forecast showed that inflation was likely to reach 6%Y by 4Q11, peak at 6.3%Y in 1Q12 and to remain sticky around 5.9% for the remainder of 2012. This forecast was based on expectations of a downwardly rigid oil price and presumably a weaker rand outlook, given that its Monetary Policy Review published a fortnight later highlighted the potential upside risks from ZAR depreciation. Further, while the MPC statement mentioned that “oil prices are not expected to decline further in the near term”, the reality is that prices fell a further US$7/bbl from US$113 at the time of the MPC meeting to US$106 in late May. The currency also strengthened from USD/ZAR6.90 to USD/ZAR6.71 in early June, although it is currently trading back at USD/ZAR6.90. And finally, unit labour costs (ULC), which are an important determinant of the SARB’s inflation forecast, are bound to have continued to ease in 1Q11 from their 4Q10 reading of 7.7%Y, especially as the big 4.8%Q upside surprise in 1Q11 GDP growth must have driven productivity-adjusted wage inflation even lower. Our initial estimates point toward ULC of some 7%Y in 1Q11.
Thus, we believe that, although the May CPI may have come in higher than the MPC’s baseline scenario, its new inflation profile that will be presented at the upcoming July 21 meeting is likely to show inflation remaining below the target band throughout 2011, and is likely to show a lower peak of some 6.1% in 1Q12, in our view.
SARB to Pull the Ace Card
This improvement in the inflation trajectory may present some credibility challenges, given that the May MPC communiqué fast-tracked the policy debate from a discussion of the relevant underlying inflation measures to one of vigilant data-watching, portending an imminent rate hike (for more details, see South Africa: Hawkish SARB Portends Earlier Tightening, May 12, 2011).
To get around the potential credibility/communication challenges, we believe that the MPC is likely to extend its inflation forecast horizon to 2013, where inflation is likely to return more meaningfully into the target band. Such a return to a ‘hump-shaped’ profile should allow the MPC to argue that the expected breach of target is mostly driven by exogenous factors that eventually fall out of the wash, thereby obviating the need for tighter money (for more details, see CPI Breaches Mid-Point: Will the SARB Pull an Ace? June 22, 2011).
Timing of Policy Normalisation?
One thing is clear: given that the SARB did not raise policy rates in May when it forecast an inflation peak of 6.3%Y by 1Q12, we believe that it is likely to consider policy normalisation only if future inflation forecasts point to a worse trajectory than that (our previous inflation forecast was in favour of the latter, hence our call for policy normalisation to commence as early as September). It could also consider raising rates if underlying inflation pressures turn out to be materially worse than expected. Our revised inflation forecasts suggest that neither of the two is likely, raising the odds that policy rates are unlikely to be raised this year. Our improved inflation profile therefore allows the SARB to ‘kick the can down the road’ – at least until 1Q12, where inflation is likely to breach the upper end of the target band.
We also believe that the SARB is likely to focus on the output gap and the prevailing rate of GDP growth to give direction on the timing and extent of policy normalisation. Global growth factors (especially in Europe) are likely to have some relevance here, given South Africa’s significant exposure to European growth conditions (see South Africa: Gauging Susceptibility to the Vagaries of European Growth, June 21, 2010). We are therefore inclined to believe that the SARB is likely to delay policy normalisation until at least 1Q12, where we look for the first of four successive 50bp rate hikes, taking terminal rates to 7.5% by September 2012. However, we must point out that, while the SARB may prefer to engineer successive, measured rate hikes, any soft patch in growth – be it local or global – could force it to pause along the way.
Source: Michael Kafe and Andrea Masia, Morgan Stanley, July 15, 2011.