AFRICA - Over- value currency

Falling FX reserves increase depreciation pressure on currencies. Sub-Saharan Africa’s (SSA) dependence on imports, due to its limited manufacturing capacity (manufactured goods make up only 13% of SSA’s GDP, but 66% of its merchandise imports), explains why its authorities prefer strong or overvalued currencies. Generally, the FX reserves of SSA oil exporters provide enough firepower to sustain their strong currencies. However, the FX reserves of oil importers (such as Kenya and Zambia) are not as large, limiting their ability to sustain overvalued currencies. Oil-importing Malawi is a case in point, having resisted calls from the IMF to devalue the kwacha (it was devalued by around 10% in August 2011, to MWK165/$1) and allow for a more flexible currency. Malawi argues that its significant dependence on imported inputs, most notably fertiliser, would mean a weak currency could significantly raise the cost of production, undermining growth. However, Malawi’s dire FX shortage indicates to us that the current kwacha/dollar exchange rate is unsustainable. Falling FX reserves increase the pressure on overvalued currencies to depreciate, and Kenya and Nigeria have seen their FX reserves come under pressure for different reasons. Kenya’s large and widening current account deficit explains the downward slide of its official reserves (see Figure 5) and +20% depreciation of the shilling YtD. Nigeria’s sharp fall in FX reserves in 2010 and the absence of a recovery in 2011, support a potential devaluation of the naira in the near term (see Figure 6).

A strong currency hurts resource-intensive countries (Nigeria and Zambia) less than significant exporters of services (Kenya) and manufactured goods. Real effective exchange rate (REER) analysis is used to determine whether a currency is misvalued, and to assess the competitiveness of a country’s exports. We have found the naira, Kenyan shilling, cedi and Zambian kwacha to be overvalued – implying they have not sufficiently depreciated to compensate for the inflation differential between the SSA countries and that of their trade partners. In theory, this suggests the respective countries’ exports are uncompetitive, as more dollars are required to purchase their exports than to purchase exports from a country with a weaker currency. However, the negative impact of an overvalued currency is blunted by the fact that SSA exports largely comprise commodities (oil, metals and minerals) that are priced in dollars. So, in our view, the export competitiveness of resource-rich countries, such as Nigeria and Zambia, is not significantly affected if their currencies are overvalued. But exporters of manufactured goods and services are more likely to be hurt by an overvalued currency. As SSA exports are light on manufactured goods (manufactured goods constitute 3% of Nigeria’s merchandise exports, 19% of Ghana’s [pre-oil] and 8% of Zambia’s) this is not a significant issue. However, Kenya has relatively high exposure to manufactured exports (37% of merchandise exports), and is a significant exporter of tourism services, so the negative impact of a stronger shilling would be more pronounced on its exports.

The KES still has a significant depreciation risk. Our REER analysis shows that the Zambian kwacha was the least overvalued, the cedi was the most overvalued, and the naira and Kenyan shilling are both significantly overvalued (Figures 1 to 4). The Nigerian authorities not only prefer a strong naira, but they have the reserves to keep it strong (but maybe not at NGN150/$1 +/-3%), even after the $10bn drop in reserves in 2010 (Figure 6). After our projected moderate near-term devaluation of the naira to around NGN155/156/$1, we think significant naira depreciation pressure could be stemmed with the current FX reserves stock of seven-to-eight months of import cover. Conversely, Kenya’s FX reserves position has been deteriorating since late-2009 (see Figure 5), falling almost 10% since they peaked at $4.1bn in August 2010; while that of Ghana has improved to about $5bn, from $2bn at YE08, implying there is more pressure on the Kenyan shilling to weaken. Kenya’s precarious FX position explains its appeal to the IMF for balance of payments support. While we note the shilling has appreciated in recent weeks, following the strong dose of monetary tightening in October, we still think there is a high risk of it weakening in the short term compared with other SSA currencies – especially if Kenya’s structural imbalances are not reversed by unwinding the expansionary fiscal programme. Our base-case scenario is for the shilling to average at KES94-95/$1 in 2012. However, we assign a 30% probability to it weakening beyond KES100/$1 in 2012. The higher risk of depreciation of the shilling, compared with a devalued naira, suggests to us that interest rates may remain higher for longer in Kenya. We are not as concerned about the Zambian kwacha, given its relatively low level of overvaluation and Zambia’s relatively sound FX position. The risk to the Zambian kwacha, in our view, is a sharp decrease in the copper price, contrary to consensus projections of stable prices in 2012. We think the improvement in Ghana’s FX reserves