Small man’s carry trade in Central and Eastern Europe: Is he really taking the bet?

Raphael Auer, Martin Brown, Andreas Fischer, Marcel Peter 29 January 2009

a

A

The strategy of borrowing (going short) in a low-interest rate currency and investing (going long) in a high-interest rate currency, the so-called carry trade, can be traced back at least to the Medici financial empire in the 15th century. While leveraged financial institutions have traditionally been the main parties in carry trades, the wisdom of this strategy is now being debated in the living rooms of many Central and Eastern European (CEE) households.

Many empirical studies highlight the risks involved in carry trading for financial institutions. The carry trading strategy is even riskier for households and small firms. Do these investors take carry trade positions only because of their limited understanding of currency and maturity mismatches? Relying on two recent studies, we argue that micro-level findings may partially alleviate policy concerns about household and small firm credit risk in the face of domestic currency depreciations. In short, many holders of foreign currency loans are in a more comfortable financial position to withstand currency shocks than is commonly believed.

Foreign currency exposure is over $350 billion

Well before the breakout of the global financial crisis in August 2007, the IMF (2005, 2006) had warned about the macroeconomic risks of strong credit growth in Central and Eastern Europe, especially as an important share of this credit was and is denominated in foreign currencies. More recent IMF analysis summarised in Rosenberg (2008) finds that the EU’s new member states in the region are increasingly exposed to currency risk, raising their exposure to the banking crisis that is currently sweeping across Europe. In particular, the difference in interest rates between domestic and euro zone or Swiss interest rates and the banking sector’s increased access to foreign funding are the main drivers of foreign currency borrowing.

It is estimated that last year over $250 billion of loans (or roughly 40% of total loans outstanding) in Central and Eastern Europe were denominated in a foreign currency (Brown et al. 2009). In addition, recent evidence from the IMF (Rosenberg and Tirpak, 2008) suggests that foreign currency borrowing is growing much faster among households than among firms. Surprisingly, in Austria, an original member of the euro area, 20% of total loans (over $100 billion) are in foreign currency. Even more startling, almost 30% of total loans to households are in foreign currency, mostly Swiss francs.

Financial regulators have expressed their fear that holders of foreign currency loans do not fully comprehend their positions and are not hedged against currency risk. A low level of financial sophistication could exacerbate the macroeconomic fallout because households possess neither the financial literacy nor the risk-bearing capacity to cope with the current global financial crisis.

Household carry traders in Austria

Beer, Ongena and Peter (2008) analyse a 2004 financial wealth survey of 2,556 Austrian households to draw a comprehensive profile of attitudes and characteristics of households involved in what one could dub “household carry trades”. Twelve percent of Austrian households report their housing loan to be denominated in foreign currency, almost all in Swiss francs. These carry traders may pose a systematic credit risk to the lending institutions should an unexpected and sharp appreciation of the Swiss franc occur, for example. Moreover, Austrian households, who are otherwise known to be conservative investors, willingly take extra risks through other novel features of Swiss franc loans, such as their variable interest rate or the indirect amortisation via a repayment vehicle that is often invested in equity.

Which households tend to borrow in Swiss francs? Table 1 lists the means of four household characteristics – income, wealth, risk aversion, and an indicator of financial literacy – that are potentially important in explaining household loan choices. It describes the average characteristics for households with (1) no loan, (2) a housing loan in local currency, (3) a housing loan in foreign currency, and (4) another type of loan. The table shows that households with a housing loan in foreign currency have the highest income, highest wealth, strongest financial interest, and lowest risk aversion of all groups. Beer, Ongena, and Peter (2008) broadly confirm these univariate findings in a multivariate setup.

Table 1. Household characteristics and choice of loan type

 



 
All households
Households with
no loan

a housing loan in local currency

a housing loan in foreign currency

another type of loan

Income (monthly, euros)

2,470
2,265
2,834
3,377
2,682
Wealth (euros)
54,666
56,461
55,448
75,126
32,726
In wealthiest 5%
0.05
0.06
0.03
0.12
0.01
Risk aversion
0.82
0.84
0.81
0.70
0.76

Interest in financial issues

0.62
0.58
0.70
0.74
0.63
Number of observations
2,556
1,622
655
89
333

 




Note: The table lists sample means for surveyed households Wealth is gross financial assets in euro. Risk aversion: 0=When I invest, a high return is more important than a lot of security, 1=Don’t agree. Interest in financial issues: 1=I like to deal with financial issues, 0=Don’t agree.

Foreign currency borrowing by small firms in transition countries

In a second micro study, Brown, Ongena and Yesin (2008) use the 2005 Business Environment and Enterprise Performance Survey of the EBRD/World Bank to examine firms' choices of loan currency denomination. Their data set covers 3,105 recent bank loans for predominantly small firms in 26 transition countries (including all Central and Eastern European countries). The study finds that firms with foreign currency cash flow are more likely to borrow in foreign currency.

Table 2 describes three indicators of firm-level foreign currency cash flow from the study: firms that export, sell to multinational firms, or are owned by foreigners. Firms with foreign currency loans are substantially more likely to access foreign currency cash flows. The authors show in their study that these firm-level results hold true in a multivariate analysis, controlling for sector and firm specific factors (e.g. firm default risk, financial transparency). Moreover, the authors further find no evidence supporting the conjecture that interest rate advantages on foreign currency funds and exchange rate expectations drive foreign currency borrowing.

Table 2. Firm characteristics and choice of loan denomination



 
All firms

Firms w/ local currency loan

Firms w/ foreign currency loan

Difference tests
Exporter
0.34
0.31
0.43
t(1,236) = 6.03
Sales to multinationals
0.18
0.17
0.24
t(1,116) = 4.16
Foreign firm
0.11
0.08
0.20
t(1,019) = 7.52
Number of observations
3,105
2,335
770

Note: The table reports sample means by choice of loan denomination. Exporter: 1=firm has export revenues, 0=otherwise. Sales to multinationals: 1=firm has domestic sales to multinational companies, 0=otherwise. Foreign firm: 1=at least 50% of ownership in foreign hands, 0=otherwise.

 

The reported difference tests are standard t-tests assuming unequal variances (Satterthwaite’s degrees of freedom in parentheses). All three tests are significant at the 1% level.
Micro factors show that aggregate risks may be smaller than feared

Policymakers have highlighted the macroeconomic risks of foreign-denominated loans for the banking sector in Central and Eastern Europe. To back their claims, they have relied on a series of cross-country studies that emphasise the importance of interest rate differentials between the funding and receiving country or institutional factors. In this column, we do not negate the prevailing macro risks in Central and Eastern Europe that are augmented by the rapid growth of foreign currency borrowings in recent years. Rather, we offer new micro evidence showing that the distribution of holders of foreign currency loans in Austria and 26 transition countries is skewed towards higher-income families and firms that have a natural hedge in their capital or revenue formation. These micro characteristics suggest that the aggregate risks in these countries may be smaller than is often feared.

The views expressed in this column do not necessarily reflect those of the Swiss National Bank.

References

Beer, Christian, Steven Ongena, and Marcel Peter, 2008. Borrowing in Foreign Currency: Austrian Households as Carry Traders, Swiss National Bank Working Paper 2008-19.

Brown, Martin, Steven Ongena, and Pinar Yesin, 2008. Foreign Currency Borrowing by Small Firms, working paper presented at the SNB-CEPR Conference: “Foreign Currency Related Risk Taking by Financial Institutions, Firms and Households”, 22/23 September 2008.

Brown, Martin, Marcel Peter, and Simon Wehrmüller, 2009. Swiss Franc Lending in Europe, Policy Paper, Swiss National Bank (available from the authors upon request).

International Monetary Fund, 2005. Emerging Europe: Is Rapid Credit Growth a Cause for Concern?, World Economic Outlook, September, 42-45.

International Monetary Fund, 2006. Household Credit Growth in Emerging Market Countries, Global Financial Stability Report, September, 46-73.

Rosenberg, Christoph, 2008. Foreign Currency Borrowing Is More Risky For Eastern Europe, IMF Survey Magazine, Vol. 37 (11), 179.

Rosenberg, Christoph B. and Marcel Tirpak, 2008. Determinants of Foreign Currency Borrowing in the New Member States of the EU, IMF Working Paper 08/173.

a

A

Topics:  International finance

Tags:  carry trade, Central and Eastern Europe, small firms, households, foreign-denominated

Principal Economist, Bank for International Settlements and Research Affiliate, CEPR

Professor of Banking, University of St. Gallen

Economic Adviser at the International Research Unit of the Swiss National Bank and CEPR Research Fellow

Senior Economist at the International Research Unit of the Swiss National Bank

Events

CEPR Policy Research