The South African Rand (ZAR) is one of the most liquid emerging market currencies in the world, but is arguably also extremely volatile, particularly in recent years. This volatility is underpinned by the sentiment-driven ebb and flow of international capital flows and tends to make forecasting the currency a notorious challenge. As such, some investors may even suggest that it is impossible or hopeless task to try and forecast the currency
This is indeed true over the short-run. The shorter the time frame, the more of a ‘random walk’ any financial market or price will tend to exhibit. In the long-run, macroeconomic fundamentals will prevail as they do for any investment. These macroeconomic fundamentals are observable and usable in making rationale investment decisions.
So which fundamentals should investors look at in terms of trying to assess the long-run fair value for a currency? Many economists tend to rely on the theory of ‘Purchasing Power Parity’ (PPP), often expressed in its most simplistic terms by the Economist’s infamous ‘Big Mac Index’, which compares the US Dollar price of Big Macs around the world.
Investopedia offers a succinct explanation of PPP
Purchasing Power Parity (PPP) is an economic theory that compares different countries’ currencies through a market ‘basket of goods’ approach. According to this concept, two currencies are in equilibrium or at par when a market basket of goods, taking into account the exchange rate, is priced the same in both countries.
In essence, if a basket of goods or any one product in one country is cheaper than that in another, the currency of the former must be undervalued. PPP is subject to significant limitations however as they often take into account goods or products that are tradable compared to those that are non-tradable. For example, even though a Big Mac in South Africa is much cheaper in US Dollar terms than most developed countries, when was the last time South Africa actually exported a Big Mac to the UK?
The central argument is that if a currency is cheap, then exports will increase and imports will decrease, thus resulting in a narrowing of the country’s trade deficit or even a trade surplus. But much like nominal wages in the real world, changes in exports and imports tend to take time as they are sticky. There are often also structural or institutional constraints that inhibit the country from taking advantage of a cheaper currency. In other words, it takes time before corporations may invest and build a new factory or mine to advantage a cheaper currency. In some cases they will not make such an investment decision no matter how cheap the currency gets due to certain structural constraints. This is quite relevant in South Africa at present, if we think about factors such as elevated annual wage increases, labour inflexibility and political uncertainty.
As such, focusing solely on measures such as PPP can often provide a distorted picture of the true value of a currency. Another metric that is similar in nature to PPP, but somewhat more useful in our opinion is the Real Effective Exchange Rate (REER). This adjusts the relative movement of the currency that you are analyzing against a single currency or basket of currencies (often the country’s major trading partners) by the difference in the inflation differential between the countries.
The South African Reserve Bank (SARB) compiles and maintains such a series and is depicted in the chart below.
The REER for the Rand has oscillated around a mean level of around 98 for the past 46 years, reaching a record high in the early 1980s around 140 (a period of high domestic inflation, which was only reflected in the currency with a lag in 1985/86) and a low near 60 during the 2001 currency crisis. As we can see, the REER only reached an average level of 70 in January 2016, when the currency traded between 16 and 17 to the US Dollar. Although this was a new nominal record for the exchange rate relative to the US Dollar, it can be argued that the Rand at this time was still not as undervalued as it was in late 2001.
The data in the above chart ends in December 2016 and since then the Rand has strengthened by a further 8%. Incorporating this further appreciation and assuming a 1% inflation differential over the last three months, it would take the current REER reading to around 92. This is not far from the long-run median of level of 98 and would suggest using this metric the Rand is at or close to fair value at present levels.
The above approaches can be refined somewhat further by looking only at a country’s terms of trade or even wages/income on a relative basis. The former looks at exports prices relative to import prices and will therefore specifically exclude non-tradable goods and products. For South Africa, where mineral exports continue to account for a majority of total exports, looking at relative changes in the terms of trade is often quite useful.
A major positive tailwind for the Rand over the past 12 months has been the relative increase in the prices for some of the country’s mineral exports relative to oil prices. Oil remains the country’s largest and most important import, given the inelasticity of demand for this product. Simply put, a country’s terms of trade will eventually be expressed via the trade balance. In our opinion, although REER is useful and certainly more informative than PPP on its own, the single most useful metric for investors to keep an eye remains the trade balance and more specifically the current account balance.
The current account balance incorporates all income flows, including service receipts and payments from tourism, as well as income receipts and payments made to non-residents holding South African assets. If a country runs a current account deficit we can say that the currency is probably overvalued and vice-versa. In general, if the deficit remains below 3% of GDP, it may not have a significant or immediate impact on the currency. As already noted, capital flows can and do tend to dominate pricing in currency markets in the short-run.
South Africa registered a current account deficit of 3.3% of GDP in 2016, narrowing from 4.4% in 2015. This was largely the result of a substantial improvement in the country’s trade balance, from a deficit of ZAR 38bn in 2015 to a surplus of ZAR 15bn in 2016.
The net deficit excluding trade and comprising only services, income and current transfers actually widened to ZAR 156bn from ZAR 140bn. For various reasons, South Africa runs a persistent income deficit in that it pays more in dividends and interest to non-residents than the country receives in income from its offshore assets. This partly reflects the large inflows into South African assets between 2002 and 2013, but also likely the difference in yields between South Africa and much of the developed world.
If we are optimistic and assume that a current account deficit of 3% of GDP represents a kind of neutral level, then we can say that as long as South Africa can run a ‘balanced’ trade account, a structural income and services deficit amounting to 3% of GDP should be manageable. However, how do we know that any specific trade balance reported in prior years is sustainable or ‘normal’? This is important because if we assume that the volume and prices for exports and imports do not change, then we can also assume that the average USD/ZAR rate prevalent in 2016 of roughly 14.70 represents a likely fair value for the currency.
One way to determine whether a current trade balance is reflective of the underlying structural trends and nature of the economy is look at total exports and imports relative to GDP. The charts below show the trend in these measures over the last 40 years.
Total exports including services as % of GDP (SARB)
Total imports including services as % of GDP (SARB)
The charts illustrate both a structural trend as well as several cyclical trends. The structural trends shows how both imports and exports (and therefore net trade) as a % of GDP increased after 1994 as sanctions were lifted and the country opened itself up. The commodity boom between 2004 and 2008 resulted in a sharp increase in both exports and imports as mineral prices including oil traded substantially higher. Nevertheless, in general we can probably say that the period since 2000 likely reflects a more normalized relationship between exports, imports and GDP following the structural adjustment post 1994.
The average ratio since 2000 for both exports and imports as a % of GDP is roughly 30%, coincidentally a level that both imports and exports reflected in 2016 and hence the return to a modest trade and services account surplus. If we assume that from a cyclical perspective these levels are ‘normal’, then we can indeed say that the average level for the currency that prevailed in 2016 is a likely fair indication of the currency’s fair value. At current levels or lower, the currency therefore appears set to become increasingly overvalued, at least using this analysis. We would expect to see a renewed widening in the country’s trade balance going forward, should these levels persist for any length of time.
What about the structural outlook?
What if a normal level for trade and services exports or imports as a % of GDP is not 30% or the average of the last 17 years? In our view, South Africa is indeed structurally not the same as it was 15 years ago. When taking into account the micro constraints that have impacted on mining investment and production, we would argue that mineral export volumes are unlikely to increase from current levels or at least significantly. This implies that a 30% level for exports relative to GDP may now actually be on the high side.
As such, if the country is to maintain a balanced trade and services account it would suggest that there is also limited scope for imports to grow relative to GDP going forward. This is particularly relevant when we consider the fact that oil prices, having averaged around $40 per barrel in 2016, may trend higher in coming years, thus increasing the ratio of imports to overall GDP. The chart below outlines how mainly mineral product imports (mainly oil) as a % of GDE (Gross Domestic Expenditure) in 2016 were in fact lower than those during 2009, when the country was in recession and oil prices were also depressed.
If the prices of South Africa’s key mineral exports do not also move higher in US Dollar terms in this type of scenario, the country’s trade balance will quickly return to a sizable deficit. At a structural level we would argue that the scope for non-mineral and services export growth is ultimately central to any kind of long-term outlook for the Rand, as well as one’s view on the future trajectory of relative mineral prices.
Total goods and services exports ( subtracting for mineral product or oil imports would amount to roughly 25% of overall GDP, assuming that 30% total exports and 5% for mineral imports is normal. Let’s assume as an example a return to $80 oil, which results in the ratio of imports to GDP increasing to 32%, a level incidentally it averaged between 2011 and 2016. Should this higher level be sustained, total goods and services exports, excluding a portion of exports to compensate for oil imports), would have to increase by roughly 8% in order for the trade account to remain balance. Assuming no change in actual export volumes, this could be easily achieved by a further 8% depreciation in the Rand and the subsequent increase in Rand-denominated export revenues.
Unfortunately it is not that easy, since the Rand-denominated value of the country’s imports will also increase. In order for the trade account to remain in balance, the total volume of imports into the country will therefore also have to decrease by 8%. In a 2012 study, the South African Reserve Bank found that the long-run elasticity of imports to demand-related price changes was around 0.8 for non-oil imports. For the sake of simplicity we can then assume a ratio of 1 and ignore oil imports from our analysis as we have accounted for them by subtracting from our total exports metric. In this case, we can then indeed say that a 8% depreciation in the Rand would lead to a 8% reduction in import volumes, excluding those of oil, which would allow the trade account to remain in balance in this scenario .
The very low economic growth and depressed domestic investment in 2016 resulted in a decline in demand for imported capital equipment. When coupled with the decline in oil prices and hence oil imports, we think it would be prudent to overweight a higher import scenario in any kind of currency analysis. As such, we would suggest that a fair value for the currency is probably somewhere between R14.00 and R 15.00 USD/ZAR taking REER into account.
If we take REER as our “lower bound” fair value for the currency, presently at between R12 and R12.50, we could argue that a fair value for the currency is in fact anywhere between R12 and R16 to the US Dollar and the Rand will tend to trade between these levels dependent upon prevailing capital flows. Ceteris paribus, assuming that the only other relevant variable to consider going forward would be the long-term differential in relative inflation rates between South Africa and the US, this fair value range would increase by roughly 4% – 5% per year if mean inflation prints at 1.5% in the US and 6% in South Africa.
As such, the recent strength in the Rand is offering investors another extremely attractive entry point to diversify into offshore investments. If we use the two most realistic ways of measuring the fair value of the currency, it is clear the Rand is now trading at or close to the lower bound of the aforementioned ranges.
This analysis can however not offer any assistance in terms of timing and should global capital flows remain supportive, the Rand could continue to appreciate in the short-run. The best strategy for most investors, as is often the case, is to ‘average’ their reallocation of capital over a period of time (perhaps 12 months) and as long the currency continues to trade near the lower bound of the discussed fair value ranges.
What would it take to align REER with a fair value based on a balanced trade and services account?
If we are correct in our assumption that there is limited scope for meaningful growth in the volume of South African exports, then only a further positive terms of trade shock, which results in a widening in the trade and services surplus to between 2% and 3% of GDP ( effectively offsetting the entrenched income deficit ) eliminating the country’s current account deficit all together, would achieve this:
Source: Statista, Blue Quadrant Capital Management
At present Iron ore prices are already trading higher than the required level, although gold and platinum remain substantially below these required levels. Should gold and platinum prices trade substantially higher from present levels, with oil remaining below $50 per barrel, it would justify the present trading range for the Rand or align the REER lower bound with a fair value consistent with a balanced current account.
The Rand following such a further positive terms of trade shock could ofcourse overshoot the long-run mean REER level of around 98. However, even during past commodity booms and periods when the current account balance was in surplus (2002), the highest the REER ever reached (with the exception of the early 1980s) was around 110.
This would suggest that even under the most optimistic of scenarios, the upside to the Rand from current levels (vs. the US Dollar) is around 11. Certainly the risk at this time is fairly asymmetric and would argue for a return to a Rand negative posture in investment portfolios.