EM currencies: are the “fragile five” really that fragile?

In biits

Five EM countries are regularly trotted out as being most at risk of a currency crisis: Brazil, Indonesia, India, Turkey and South Africa.

Their currencies fell by between 13 per cent and 23 per cent between May and August this year (before recovering somewhat in most cases).

But a new index from Nomura makes a mockery of the distinction between the Fragile Five Biits – as the five have been dubbed – and the rest.

According to Nomura, Ukraine and Serbia have a greater chance of a currency crisis. The Biits are not similar at all, it seems. Turkey is much more at risk of a plunge in its currency than Brazil and Indonesia, which are at the opposite, safer end of the scale. Does it make sense to group the five together at all?

Source: Nomura. Click to enlarge

Readers may be tempted to call the index a pile of nonsense. After all, the Indonesian rupiah dropped the furthest of the five after May 13. But the research should not be dismissed out-of-hand. Analyst Peter Montalto of Nomura says the index will not predict which emerging markets investors will choose to pull out of, but it does cast light on the underlying fragility of each one.

It helps to understand how the index is compiled. Nomura’s global emerging market risk index (GEMaRI) is intended to gauge the risk of a currency crisis, defined as a movement in the exchange rate of three standard deviations from the two-year mean.

The scoring is based on 16 subindices, each weighted by its role in a currency crisis in the past. Big stock market falls and a rapid deterioration in the current account deficit are given the greatest weight. Note that the indicators are not chosen arbitrarily but based on cross-country data going back to 1997.

Four of the other big signs of a looming currency crisis, ranked by importance, are: a hike in the real interest rate of more than 2 percentage points over the quarter; a fall in forex reserves relative to short-term external debt; a current account deficit larger than 3.5 per cent of GDP; and a forex-to-imports ratio of less than 4.

Indonesia rings two of these four alarm bells. The ratio of its foreign exchange to short-term external debt fell from 2.8 to 2.6 between 2011 and 2012, and is forecast to fall further. Its current account deficit in Q2 was equal to 4.4 per cent of GDP.

But more importantly, according to Nomura, the two loudest sirens have not sounded. From Q2 to Q3 the Jakarta stock exchange did not drop by more than a tenth, and year-on-year the current account deficit has not fallen by more than 3 percentage points.

Source: Bloomberg

Nomura’s number-crunching is welcome. It is interesting that a ballooning trade deficit matters more than the size of the deficit. A hike in real interest rates is more likely to herald a crisis than a diminishing chest of forex reserves, if we believe the research. It also seems east Asia has learnt from its 1997/98 crisis, as the World Bank recently concluded.

But the index has its limitations. For one, it is too binary. The Jakarta stock exchange index fell by about a fifth from its height in May to its low in August. But from Q1 to Q2 and from Q2 to Q3, as measured by Nomura, the respective falls were less than a tenth. A difference in the measuring period has a big impact, due to the weight given to stock prices in the index.

This may explain why the rupiah’s fall seems less significant, seen through the lens of the Nomura index.

Secondly, and perhaps more crucially, crises are partly self-fulfilling. A sell-off can trigger a decline in the currency, which can cause yet more sell-offs as investors lose confidence in the country. Sell-offs can be based on blind panic, as well as a clear-headed analysis of a country’s fundamentals. So to some extent, country rankings may prove irrelevant.

Related reading:
Asia: learning the lessons of 1997, bb
Does renewed acronym anxiety spell crisis? Comment, John Authers
Who’s afraid of the big bad capital withdrawal? bb