By James Lord, Michael Kafe and Andrea Masia of Morgan Stanley.
In South Africa, it is no secret that the currency’s phenomenal strength has been a source of concern – and, occasionally, even frustration – for the authorities. In real effective terms, the rand has rallied by around 40% since the recent low at the end of 2008. Along with BRL, the rand has seen the biggest rise in REER terms since EM currencies started to recover (in general) from the lows seen at the end of 2008 following the collapse of Lehman Brothers.
In addition to the rapid appreciation in the value of the ZAR, the current level of the REER may be causing some concern too. The ZAR is now some way above the 10y average, and with some uncertainty regarding the strength of aggregate demand in the global economy, this rapid strengthening of the currency in recent months has fuelled the currency intervention debate. It is unsurprising, therefore, that a recently released ANC discussion document, which will be tabled at its National General Council meeting next month, lists a “more competitive exchange rate” as a key policy objective for the future. This publication prompted a minor wobble in the value of the currency, though it quickly regained its poise.
While a lot of debate has taken place, on domestic and international platforms alike, about how to sustain a more competitive (read: weaker) rand, it is not too surprising that a permanent solution remains to be found – not least because the equilibrium level of the rand is itself a function of dynamic macroeconomic relationships, and hence is a moving target that cannot be easily pinned down and calibrated. For the record, our fair value estimate for USDZAR, based on our proprietary variant of the monetary approach, is 8.40 – i.e., an overvaluation of some 15% from the current level of 7.25.
The fact that South Africa, a relatively open economy with a trade intensity ratio of some 60% – up from less than 40% in the early 1990s – has adopted a flexible but credible inflation-targeting framework makes it fundamentally difficult – if not impossible – to contemporaneously target the currency without compromising monetary independence. But this does not mean that the authorities are completely hamstrung: Although the SARB has adopted a hands-off approach to currency intervention – partly because of the huge hole that it dug in its balance sheet while trying to defend the currency in 1998 (see “A Survey of FX Intervention Policy”, CEEMEA Macro Monitor, April 26, 2010) – it nevertheless has an FX accumulation strategy that seeks to cream off excess liquidity during periods of undue pressure on the rand.
Following the Treasury’s stated intention earlier this year to partner with the SARB and intensify their efforts to lean further against the wind during periods of rapid capital inflows, the foreign exchange authorities have no doubt stepped up their opportunistic accumulation of foreign exchange. The foreign exchange position of the SARB at the end of July showed an increase of some US$1.1 billion in FX holdings. We estimate that approximately US$574 million of that was due to valuation effects, while the remainder (US$533 million) was due to naked FX purchases. A US$480 million increase in government deposits suggests that most of this (a little under 90%) was executed on behalf of the National Treasury. This brings total naked FX accumulation by the SARB/National Treasury to at least US$2.4 billion for the year to date, on our estimates.
The most obvious constraint with this strategy, however, is that it is an expensive exercise (the associated sterilisation costs are prohibitive), and does not seem to have had a major impact on the exchange rate – although of course the counterfactual is virtually impossible to prove. It is also important to realise that the real issue for South Africa is not so much the authorities’ ability to engineer currency weakness by aggressively purchasing foreign exchange. In our opinion, the real challenge is how to defend that desired level of the currency, once it is attained. In a market where daily turnover can reach levels as high as US$14 billion, against gross foreign exchange reserve holdings of US$42 billion (i.e., just three days of market turnover), the hard truth is that the SARB simply does not have the resources to defend/manage the currency if it were to depreciate beyond its desired equilibrium level. One must remember that, since the 1990s, the SARB has depended on borrowed reserves, and it was not until June this year that it fully unwound all of its borrowed FX holdings. Surely, it will not want to run the risk of being forced to return to creditors anytime soon?
So What Other Options Do the Authorities Have?
Reintroduction of the Dual Rate Exchange Rate
South Africa is no novice when it comes to capital controls, which existed as early as 1961 in this economy. The most prominent period of controls in the country’s history, however, dates from September 1985, when the authorities declared a debt standstill and adopted a dual exchange rate system, which had separate markets for current account transactions (commercial rand) and capital account – mostly portfolio capital – transactions (financial rand). The purpose of this dual rate system was to prevent large amounts of capital from leaving the country at the time, and to help insulate the commercial rand market from the inevitable consequences of extremely volatile flows on the capital account. Empirical evidence suggests that this strategy was indeed successful, in that currency volatility declined during this period.
With the introduction of a democratic government in 1994, the economy gradually opened up and the dual rate system was finally unified into a single currency in March 1995. Since then, South Africa has earned a solid reputation for its commitment to gradually unwinding all existing foreign exchange controls. In our opinion, to reintroduce any form of exchange controls would be a step backwards for the country, and could well engender market uncertainty. Actually, our preferred option would be for the authorities to rather consider removing the remaining restrictions on residents.
Economic literature on the effectiveness of capital controls is inconclusive. A recent IMF report highlights that prior attempts to impose restrictions on the flow of capital, while well intentioned, have had mixed results. By and large, capital controls have had little impact on the volume of capital inflows. Accordingly, any change in currency performance following the imposition of such controls should not be attributed to a change in the flow of capital resulting from the more restrictive capital account.
Capital Inflow Tax
As indicated earlier, the ruling ANC put forward a discussion document last week which mooted the imposition of a tax on short-term capital inflows. This is one of the many instrumental variables that we believe technocrats at the SARB and Treasury have already debated at length and have not found motivation to implement (yet?) – not least because it appears to have failed elsewhere (e.g., in Brazil when it was introduced to fixed income flows in March 2008, lifted in October 2008, and then re-introduced and expanded one year later to include equity flows too).
Although it is true that the pace of BRL gains began to slow following this reform, there is little evidence that the introduction of the tax caused it. Indeed, EM currencies in general tracked sideways from around this period onwards, and the relative performance of Brazil is not much different to that seen prior to the introduction of the tax. What’s more, the tax did not have much of an impact on the level of net private financial inflows either. Following a minor fall in inflows during the subsequent month, inflows quickly rose again.
We suspect that something similar would happen in South Africa. A tax of 0.5-1.0% is unlikely to have much of an impact on the level of capital coming in, beyond any knee-jerk reaction. Accordingly, any weakness in the value of the rand following the introduction of this sort of tax ought to be viewed as a tactical buying opportunity, provided global sentiment towards risky assets remained benign.
Although an important appeal of this tax proposal is the fact that it could help boost the fiscus (as opposed to the presently expensive strategy of opportunistic reserve accumulation), our understanding is that the authorities fully appreciate the importance of portfolio flows to the funding of South Africa’s structural current account gap, and are perhaps of the opinion that the potential risks of a permanent exodus of vital capital probably outweigh the minimal fiscal benefit that such a tax proposal offers – particularly bearing in mind that it has in fact failed elsewhere recently.
We also note that the tax proposal was mooted in an ANC discussion document. As the Finance Minister is a member of the party himself, we doubt that he would front-run the party faithful and implement a party draft proposal before it even has the opportunity of going through the full process of discussion and possible adoption at the party’s 2012 national economic policy conference. We note that this is still a long way off – and by the time it is up for adoption, the macro backdrop may have changed. If we are wrong, however, and the authorities do implement this as early as in the October 2010 or February 2011 Budget, our opinion remains that they are likely to tread cautiously and levy the tax at a relatively benign rate that serves much more to signal intent than to penalise foreign capital.
Holding Period Restraints
Third, the authorities could consider imposing a holding period (e.g., one year) on portfolio investments. Technically, this should help to discourage fast money investors. The risk with this strategy of course is that there could be a sudden stop in portfolio inflows – with potentially disastrous implications for the currency – as few international investors (especially in fixed income space) would be willing to lock in a one-year view on a currency as volatile as the rand. Given the recent experience in Nigeria, we believe that even equity investors – who are arguably the more sympathetic of the two – could become gun-shy very quickly.
Capital Reserve Requirements
Finally, the authorities could require non-residents to place a certain proportion of notional purchases of their South African securities in non-interest-earning deposits at the SARB for a period of time. This appears to have worked in Peru, as part of a broader set of restrictions, but did not have a great success rate in Colombia. Again, this would signal a step backwards for South Africa, in our opinion, and would be very difficult to monitor. We do admit, however, that, if the entry rate is more indicative than punitive, it could well turn out to be a short-term panacea.
FX Intervention: Pushing on a String?
Even so, one must remember that the currency’s near-term strength may also have to do with the announcement that NTT has made a £2.1 billion offer for Dimension Data – the country’s largest IT provider (this deal is roughly the size of the entire 1Q10 current account deficit). In the past, the currency has rallied significantly each time a chunky FDI transaction has been announced. Information on the timing of the flow, and any off-market tranches that the SARB may want to take up directly from the buyer, are unavailable at this stage. We also note that, as Dimension Data is close to 50% foreign-owned, not all the FX flow will be destined for South Africa. We maintain our view that USDZAR closes the year at 7.80, and that, given how hard the currency has rallied, the risk/reward is now titled towards buying USDZAR over the next few weeks at least.
Source: Morgan Stanley, August 12, 2010.
