What Drives FX Borrowing in Eastern Europe?

15 percent of outstanding private sector credit in Eastern Europe is now denominated in (or indexed to) foreign currencies—compared with only four percent a decade ago. The private non-financial sector, in particular, is increasingly exposed to currency risk, raising alarm among those who are concerned with financial stability. Moreover, Euroization has accelerated in the years since these countries joined the EU.

Why and how does EU membership induce borrowers and lenders to switch to euros and Swiss Francs? As I discuss in a recent paper with Marcel Tirpak , the answer is not straightforward. As shown in Figure 1, there are striking differences between individual new member states who all joined the EU in 2004. A cross-country analysis can provide some interesting insight about the interaction of economic incentives and country-specific characteristics. It also highlights how EU membership is fundamentally changing the policy-making environment in the new member states.

Several hypotheses have been offered on what drives East European households and corporates to shy away from loans in zlotys, lats and forints. The most obvious is that borrowers are simply ducking high domestic interest rates. But, as my IMF colleague Olivier Jeanne (2003) points out, these borrowers–if they are rational–will weigh the lower cost against the risk of a devaluation. The currency regime cannot alone explain borrowing behavior (intermediate Group C in Figure 1 contains both countries with fixed and flexible regimes), although the outliers suggest that countries with fixed exchange rate regimes (Group A) may be more prone to taking the currency risk than those with flexible regimes (Group B). Recent research at the ECB (Basso et al, 2007) stresses supply side factors—the availability of funding through the presence of foreign banks. Yet others suggest that imminent euro adoption, foreshadowed by EU or ERM membership, will induce both borrowers and lenders to take greater currency risk. Finally, one would expect that a country’s economic policies, especially regulatory measures to curb such lending, will play a role.

Our paper tests these and other possible factors in a panel regression model that uses quarterly data for 1997-2007 and includes nine new EU member states from Eastern Europe plus Croatia. A surprisingly limited number of factors hold up to statistical scrutiny. In fact, we found only four:

· The difference in interest rates between domestic and eurozone interest rates drives foreign currency borrowing, as suggested by economic theory. But unlike in other parts of the world (e.g., Latin America), past exchange rate volatility has no statistically significant effect. One reason may be that EU membership increases economic agents’ willingness to assume currency risk. If anything, people expect their currency to further appreciate as these countries converge towards Western European price and income levels. The currency regime per se or ERM2 membership have no measurable impact in our model.

· The banking sector’s dependence on foreign capital, as measured by the loan-to-deposit ratio is a strong contributor to foreign currency borrowing. Banks refinance themselves abroad and the currency risk is passed to their clients, if only because they are often not allowed to hold open currency positions. Interestingly, we find that it is irrelevant whether such foreign funding is channeled through domestic banks borrowing abroad (e.g., through syndicated loans) or foreign-own banks drawing on credit lines from their parent banks. Indeed, a few foreign banks have been reluctant to get into foreign currency lending because they burned their fingers in the Argentinean currency board collapse or the ERM crisis.

· Openness, captured by the relative size of foreign trade, matters. Revenues from abroad make it easier for corporates to hedge their foreign currency exposure. This does not seem to be the case, however, for households; remittance flows are not found to increase foreign currency borrowing.

· Regulatory policies (e.g., higher risk weights for foreign currency loans), which we model through an index measuring their severity, have some measurable effect but it is pretty weak. Their impact disappears entirely if direct borrowing from abroad is included in the dependent variable. This illustrates that regulations imposed by domestic supervisors are largely ineffective if borrowers have direct access to lenders abroad, as is the case within the EU.

While country-specific factors play a role, foreign currency borrowing turns out to be a by-product of EU membership. This appears to work through various channels. First, by fully liberalizing the capital account, the EU offers borrowers increased access to foreign funding, both through domestic banks affiliated with foreign parents and directly from abroad. Secondly, by increasing trade openness, it provides hedging opportunities, especially for the corporate sector. Finally, EU membership seems to boost the private sector’s confidence in exchange rate stability and imminent euro adoption, making it consider a devaluation a low-probability event.

Policies to steer borrowers and lenders away from currency risks—at least until countries eventually adopt the euro and mismatches largely disappear—need to adapt to the new playing field created by EU membership. As mentioned above, regulatory measures have limited effect, at least for corporate lending, because access to foreign financing directly from abroad makes it easy to circumvent them. Since under EU law capital account restrictions are not an option, any measure to address foreign currency exposures will therefore require an ever closer cooperation between supervisors in home and host countries. At home, this needs to be supplemented by rigorous stress testing of banks, appropriate guidelines for foreign currency management and efforts to educate consumers about the risks involved.