Currency markets are an important source of risk premia. Understanding why some currencies offer high returns and whether these returns provide adequate compensation for any additional risk they carry is of critical importance to all international investors.
But which currencies offer high returns? Traditionally, the answer was easy – ‘high interest rate currencies’. For decades, an investor who entered a ‘carry trade’ by buying high interest rate currencies and funding the position using low interest rate currencies was particularly successful. This ‘naïve’ strategy has generated high returns over the last 35 years, with better risk-return properties than a simple buy-and-hold strategy in the US stock market.
But the carry trade is akin to picking up pennies in front of a steamroller – unexpected exchange rate adjustments can quickly turn the slow accumulation of gains into losses. In fact, the carry trade is known to perform poorly in volatile markets and experience large return drawdowns (Menkhoff et al. 2012). This characteristic is consistent with the standard financial notion of risk, i.e. low returns in bad economic times. Indeed, Lustig et al. (2011) find strong evidence that the currency carry trade reflects a form of ‘global’ risk, different from individual country-level risk.
This 'global risk' is found to explain the average returns across currency portfolios, appearing to complete this traditional view of the currency market; high returns are available on high interest rate currencies providing compensation for bearing carry trade risk. But recent findings show this traditional currency market view of the risk-return trade-off is only part of a bigger picture.
In May 2013, the then US Fed Chairman Ben Bernanke sparked massive fluctuations in the foreign exchange market when he announced the end to US quantitative easing (QE). If the traditional view of currency risk was exhaustive, then the resulting ‘taper tantrum’ should have been particularly bad news for high interest rate currencies.
Instead, the popular press cast light on countries’ current account positions. The current account broadly captures the amount countries borrow from the rest of the world in a given period. Indeed, Alice Ross, writing for the Financial Times, noted that the Indian rupee’s sell-off during the summer of 2013 was related to it being “one of the most vulnerable emerging market currencies due to its current account deficit.”
Moreover, while the Australian dollar – a well-known ‘carry currency’ – depreciated against the US dollar by around 16% between May and September 2013, many emerging market currencies, which appeared substantially riskier from an interest rate perspective, experienced much smaller depreciations. The Chilean peso, for example, offered an interest rate over twice as large as that in Australia and yet experienced a depreciation less than half the size of the Australian dollar. In South Korea, where the interest rate was comparable to Australia, the Korean won depreciated by only 1% against the US dollar.
A differentiating factor between Australia and these emerging market countries is the external account. The external account can be thought of as the accumulation of borrowings over time.
In Australia’s case, this build-up of borrowings had resulted in an external debt position equal to almost 60% of GDP at the point of Bernanke’s QE comments. In contrast, while the emerging market countries were also debtors to the world, the positions were considerably smaller: 10% of GDP in Chile and 7% in Korea.
This anecdotal evidence suggests a second channel related to countries’ holding of external assets and liabilities might be driving currency risk beyond the interest rate channel.
The link between currencies and global imbalances
In a new research paper (Della Corte et al. 2016), we provide formal evidence in support of this anecdotal story. The extent to which countries borrow on average varies significantly around the world – hence the term ‘global imbalances’. Within this variation emerges a clear finding: net debtor countries issue the riskiest currencies.
Stepping back, for the anecdotal evidence to be supported, we should expect the world’s largest debtors to offer higher returns as compensation for the extra risk they contain.
But there is a further twist. Countries do not always issue external debt in their own currency. Many countries are, in fact, forced to issue debt in foreign currencies. In emerging market space almost all countries issue the entirety of their external debt in US dollars, euros or Japanese yen. This phenomenon, known as ‘original sin’ in the economics literature (e.g. Eichengreen and Hausmann 2005), has additional implications for risk.
If a country issues debt in its home currency, then a depreciation of the currency is good from a valuation perspective – the debt is worth less and hence the liability is lower. In contrast, if the debt is issued in US dollars, a depreciation of the local currency against the US dollar is bad – it increases the value of debt, worsening the external debt position.
In the empirical analysis, we therefore form currency portfolios sorted on the basis of countries’ net foreign asset positions (foreign assets minus foreign liabilities) and the proportion of foreign debt issued in foreign currency.
We find a wide spread in returns. A portfolio of countries with a combination of a large external debt position (relative to the size of the overall economy), with the majority of debt issued in foreign currency, generated the highest average currency return over our 30-year sample. In this large sample of developed and emerging market currencies we find this extreme portfolio generated a return of around 5.3% per year (after transaction costs), relative to a return of only around 0.9% on a portfolio of creditor currencies issuing debt in local currency.
A currency market investor can therefore enter a zero-cost ‘global imbalance’ strategy that buys debtor currencies with most of the external debt in foreign currency and funds the position out of creditor currencies with most of the external debt in domestic currency, and expect to earn a positive currency return. We find that the investment performance of the strategy is impressive. The Sharpe ratio (a measure of the return relative to risk) is found to be comparable even with the currency carry trade.
But perhaps the finding simply indicates the trade has a strong relationship with the currency carry trade.
A new source of currency market risk?
We explore this possibility by testing if the returns to the currency carry trade and the global imbalance trade are the same. We find that while the two trades are related, a big difference remains – there is potentially significant economic information contained within the global imbalance trade not contained in the returns to the currency carry trade.
We test for the presence of economic content, using the returns to the global imbalance trade as a proxy for a novel global imbalance ‘risk factor’, to assess if the factor can explain the average returns to a large cross-section of currency portfolios.
If creditor currencies earn high returns in ‘bad times’ it suggests they offer insurance against falls in investors' consumption, and hence require a higher price. It also implies that currencies which correlate more with this risk factor are the ones more exposed to global imbalance risk and thus require a higher return premium.
In empirical asset pricing tests, we show that this global imbalance risk factor does provide significant explanatory power to describe average currency returns and hence contains economic content important to currency market investors, as well as to anyone whose foreign portfolio indirectly exposes them to currency market risk.
Importantly, the risk factor provides economic content beyond the economic content contained in the risk factors previously documented by Lustig et al. (2011) to explain average currency returns.
But this finding raises a question: what exactly is this ‘global imbalance’ risk and why should debtor currencies depreciate during bad times?
The mechanism: Why are debtor countries risky?
Recent theoretical developments in currency markets provide the answer and complete this risk-based story.
Gabaix and Maggiori (2015) develop a modern portfolio-balance theory in which a financial intermediary holds currencies. Risk arises from the intermediary having a mismatch in the currencies in their portfolio. The intermediary’s capacity to bear risk fluctuates through time. In volatile periods – such as following the collapse of Lehman Brothers – this risk-bearing capacity falls dramatically and holding a large quantity of any currency becomes particularly burdensome.
But which currencies is the intermediary likely to hold on their balance sheet? It turns out that debtor currencies feature prominently. Households in the debtor country borrow in foreign currency to buy goods and services from firms in the creditor country. The intermediary is therefore forced to hold an excess of debtor countries’ currency. To compensate, the intermediary expects a positive currency return on any currency mismatch.
The theory also provides additional important implications. Over time, when risk-bearing capacity falls, debtor currencies should depreciate by the most because the intermediary will instantly require a larger expected return to hold these currencies.
We test this implication in the data and document strong empirical support for the prediction. When foreign exchange volatility rises, debtor currencies depreciate by significantly more than creditor currencies. Interestingly, this finding continues to hold even when controlling for interest rates. In fact, when volatility rises, the interest rate is significantly less important than the external account of the country for determining the impact on exchange rate fluctuations.
Global imbalances are not a new topic, and indeed economists have discussed their sustainability for years (e.g. Mann 2002). But the link to financial markets and asset prices has been relatively unexplored. While reference is frequently made to this possible link in the financial press, we document rigorous evidence linking currency returns to the global imbalance phenomenon.
The key finding is that debtor countries issue riskier currencies – a risk that is amplified if the debt is issued in foreign currency. This ‘global imbalance risk’ is novel and not subsumed by other measures of currency market risk. The finding also has strong theoretical foundations; financial intermediaries require a premium to compensate them for bearing currency mismatches on their balance sheet, a mismatch that is largely driven by debtor currencies in an imbalanced financial world.
Della Corte, P, S J Riddiough, and L Sarno (2016), “Currency Premia and Global Imbalances,” CEPR Discussion Paper No. 11129, Review of Financial Studies, forthcoming.
Eichengreen, B and R Hausmann (2005), Other People’s Money – Debt Denomination and Financial Instability in Emerging Market Economies, University of Chicago Press: Chicago and London.
Gabaix, X, and M Maggiori (2015), “International Liquidity and Exchange Rate Dynamics,” Quarterly Journal of Economics 130, 1369-1420.
Lustig, H, N Roussanov, and A Verdelhan (2011), “Common Risk Factors in Currency Markets,” Review of Financial Studies 24, 3731-3777.
Mann, C L (2002), “Perspectives on the U.S Current Account Deficit and Sustainability,” Journal of Economic Perspectives 16, 131-152.
Menkhoff, L, L Sarno, M Schmeling, and A Schrimpf (2012), “Carry Trades and Global FX Volatility,” Journal of Finance 67, 681-718.