Picture: 123RF/Maxim Kazmin
Picture: 123RF/Maxim Kazmin

SA has a substantial sovereign debt problem. Much has been made, rightly, about the revaluation of the government’s debt to junk status, after Moody’s joined S&P and Fitch in downgrading the country in March.

But one dimension of this saga that has not received proper attention is the responsibility that corporate SA has in the government’s debt problem.

To be sure, the issues which commentators highlight when discussing SA’s sovereign debt, like perennially low growth, are important. However, the composition of corporate liabilities impose a social cost that ought to make us rethink the debate on free capital flows.

To understand why, consider the "original sin" in international debt financing for an emerging market: borrowing a large share of your external debt in foreign currency. As we saw in the devastating 1998 Asian crisis, having a currency mismatch between your tax revenues (which you collect in local currency) and your debt obligations (which you owe in foreign currency) only exacerbates downturns.

It creates a vicious spiral. As your currency depreciates, it leads to a higher risk of default, which, in turn, increases the cost of servicing debt, further undermining the country’s creditworthiness.

But the fact that SA now faces a sovereign debt crisis is odd. The country has actually made huge strides in limiting its exposure to foreign currency debt. Consider that in 2007, before the financial crisis, roughly 60% of external government borrowing was in foreign currency. Recent data shows that now only about 20% of external debt is in foreign currency. This is an unambiguous improvement — the technocrats at the Reserve Bank and ministry of finance should be lauded for it.

Now consider the implications: in principle, the government could simply "print away" its local currency debt liabilities (inflation risks notwithstanding). And, since its foreign currency obligations are small, the default risk (and cost of servicing its debt) shouldn’t be significant.

Why, then, has SA’s sovereign debt been relegated to junk status? Why have yields on credit default swaps (CDSs) for SA bonds, which are a measure of default risk, more than doubled in the past two months?

One way to get a handle on this puzzle, and thereby understand the underlying dynamics, is to decompose the price (known as yield "spreads") which SA pays on its sovereign liabilities. Wenxin Du and Jesse Schreger, professors at the University of Chicago and Columbia University respectively, offer an innovative way to do this. They suggest we decompose the spread that an emerging-market government pays relative to the US government into, first, the outright default risk and second, the currency risk.

First, for the outright default risk, we create a financial instrument through derivatives markets that replicates the cost of borrowing in local currency for a riskless borrower and compare our actual cost of borrowing in rands to that benchmark. (Technically, we do this by comparing SA’s actual sovereign yield curve to the US yield curve, a riskless benchmark, swapped into rands using cross-currency swaps.)

This gives us a benchmark for what SA would pay on its debt if the probability of default was zero. From this, we can impute the market-implied currency risk, by comparing US nominal yields to SA yields, differencing out the implied outright default risk.

The result can be seen in the graph shown here, which shows the spread between US and SA government bonds, the CDS yield, the market-implied risk of local currency default, and the currency risk.

The result is remarkable. It shows that the spread due to outright default risk is near zero between 2010 and this year. For context, Du & Schreger’s study of 14 other emerging markets reports an average spread of 150 basis points; for SA, the average in the past decade is 25 basis points.

What does this mean? Why does it matter?

It shows that SA pays anomalously low local currency spreads, so the cost of borrowing is probably driven by the foreign currency debt.

But how do we reconcile the fact that the market is pricing in a high risk of default, when this clearly isn’t coming from our local currency debt while our foreign currency liabilities are small?

One answer clearly recommends itself: the market is most likely pricing in the government’s exposure to corporate liabilities.

Data on the currency composition of consolidated corporate liabilities is notoriously difficult to get. Beyond a company’s obvious desire to keep that sensitive information private, off-balance-sheet items like over-the-counter derivatives and tax-haven financing make accurate attribution difficult.

Maybe the best data we have, again from Du and Schreger, suggests that earlier in the past decade about 70% of SA’s corporate external debt obligations were denominated in foreign currency. There is some evidence that this number has probably stayed stable.

With that level of foreign currency exposure, SA’s corporate sector is vulnerable to the ramifications of the original sin. So it appears that the market is pricing in the government’s reluctance to print more money to diminish the corporate sector’s exposure, as well as the consequences for the country of this corporate debt, should the rand fall.

This is a classic case for regulation. And I’m not the first to notice it: academics and policymakers are beginning to reconsider the case for free capital flows in light of this general pattern.

This mismatch in liabilities within SA’s corporate sector also has consequences for households, since higher government borrowing costs mean the state has limited resources to finance its priorities in times of crisis, like now.

Of course, rehashing the debate on free capital flows must wait for another day.

But, given the pattern of facts here, SA would do well to revisit the merits of capital controls as a macroprudential tool. It matters for our collective economic good — both in this crisis and the ones to come.

  • Mashwama is studying for a PhD in economics at Harvard University