Should capital flows in the eurozone be controlled?

Summary:

– Liquidity trap situations make the use of capital controls effective in restoring the effectiveness of monetary policy

– Devereux and Yetman find that capital controls are nevertheless undesirable because they hinder risk diversification

– In a monetary union, however, they can mitigate the recession for peripheral countries in the event of an increase in risk premiums

– They can also be considered as a macroprudential policy, especially since they have no direct negative effects on partner countries, unlike competitive devaluation policies.

Last March, the Central Bank of Cyprus introduced capital controls to complement the crisis exit measures negotiated with the European Union. Intended to prevent capital flight following the decision to tax deposits, these controls are still in place today. However, they run counter to the capital account liberalization measures required for all eurozone countries. [1] What does economic theory tell us about the effectiveness of capital controls? Should they be allowed within a monetary union, particularly in times of crisis?

The monetary policy trilemma

Mundell’s triangle—or triangle of incompatibility—is an economic principle that has been known since the 1960s. [2] It states that an open economy cannot simultaneously achieve the following three objectives: a fixed exchange rate regime, an autonomous monetary policy (i.e., the free determination of its key interest rate), and the free movement of capital with other countries. Thus, each country or economic zone finds itself in a trilemma that consists of choosing two of the three objectives and abandoning the third. Each of the eurozone countries has thus given up autonomous monetary policy by adopting the single currency and the free movement of capital within the zone. Elsewhere, many emerging economies, including China, favor capital controls in order to maintain an independent monetary policy while opting for an exchange rate pegged to the currency of an advanced economy, usually the United States. These controls can take many different forms, ranging from incentives (taxes) to coercive measures, with the aim of limiting capital flows to or from abroad.

While choosing one of the three corners of the triangle over another is already a difficult calculation in « normal » times, financial crises significantly change the situation. In particular, liquidity trap situations, such as those we have experienced since the 2007-2008 financial crisis, add an extra dimension to the trilemma: once key interest rates have been lowered to zero, they effectively render traditional monetary policy unusable. This is demonstrated by Michael Devereux and James Yetman in a recent working paper. [3]Economies that have adopted a flexible exchange rate, such as the euro zone vis-à-vis the United States, find themselves doubly penalized in that they are both affected by external shocks—especially since the free movement of capital promotes financial contagion—and, moreover, lack monetary autonomy. Thus, the European Central Bank gradually reduced its rate to zero following the US financial shocks (the collapse of the subprime market in 2007, then the bankruptcy of Lehman Brothers in 2008).

Should we therefore consider the temporary introduction of capital controls, both in the event of an external financial shock – in order to limit its transmission to the monetary union – and/or in the event of a crisis (banking or sovereign debt) within the zone? In the longer term, can they constitute an effective macroprudential policy?

A threat to risk diversification

The choices made by international investors depend on the interest rate differential between countries, as well as between different investments within a given country. Introducing capital controls makes it possible to limit inflows and outflows while allowing a non-zero interest rate differential between the domestic economy and the outside world to be maintained. In the event of a liquidity trap abroad, the national economy’s key interest rate remains independent and can be set according to internal objectives, such as growth and price stability. Monetary policy is therefore effective and the recessionary effects caused by external shocks are limited.

However, Devereux and Yetman find that capital controls are undesirable under a flexible exchange rate regime because they reduce the level of economic « well-being » in both normal and crisis times. They create a loss of economic efficiency by hindering risk diversification and the optimal allocation of resources to the most productive investments. Although they can help the economy escape from a temporary liquidity trap, capital controls are therefore not desirable as a means of protecting against external shocks because the losses associated with risk diversification for investors would always outweigh the gains in terms of monetary policy autonomy. Given a certain degree of capital mobility, the freedom to raise the key interest rate when other economies are forced to lower theirs can certainly pay off (by attracting liquidity, for example). But this measure is « suboptimal » because it is made possible by the presence of capital controls at a given moment in time without, however, generating sufficient gains to justify its introduction.

Potential gains within a monetary union

Another recent study by Emmanuel Farhi and Iván Werning shows that capital controls may, on the other hand, be desirable within the zone, particularly during temporary shocks to risk premiums. [4] In the early 2000s, the countries on the « periphery » of the euro area, notably Spain and Greece, experienced massive capital inflows, productivity gains, and increases in nominal and real wages, before massive capital outflows and an explosion in risk premiums beginning in 2010. These countries could theoretically benefit from temporary taxes on foreign capital. Why?

Capital controls are in reality an imperfect instrument for manipulating terms of trade. In a monetary union, countries by definition lose the ability to vary their currency exchange rates as an instrument of economic policy. However, by allowing a country to maintain a real interest rate that differs from that in the rest of the zone, capital controls alter the intertemporal choices of economic agents. Consider, for example, that risk premiums increase in peripheral countries. The cost of credit increases, and production and consumption fall below their optimal levels. A tax on capital outflows, or equivalently a subsidy on international borrowing, would lower the nominal interest rate (by keeping the supply of capital artificially high) and thus encourage consumption and investment decisions to be brought forward in time.

Although eurozone treaties currently prohibit their use, capital controls would therefore be a particularly effective countercyclical instrument for reducing the impact of temporary variations in risk premiums. According to Stephanie Schmitt-Grohé and Martín Uribe, they would have significantly reduced unemployment and sovereign debt in the peripheral countries of the eurozone. [5] In Cyprus and Iceland, where capital controls were effectively implemented to directly mitigate the collapse of the banking sector, the macroeconomic gains beyond the restructuring of the banking system itself remain little known to date.

Financial frictions and macroprudential policy

More generally, the presence of financial frictions also tends to make capital controls a desirable economic policy in « normal » times (outside periods of crisis). Collateral constraints in particular lead to excessive borrowing during periods of economic boom (or bubble). They make borrowing capacity proportional to the value of collateral, even though the latter is overvalued during boom periods. Excessive borrowing in turn increases the risk of a crisis occurring, as well as the scale of any potential crisis. During a crisis,fire sales of collateral to reduce debt wipe out the value of these assets, further aggravating the crisis. The negative externality is all the greater in an open economy because the freedom to exchange capital with other countries amplifies excessive borrowing in normal times.

Thus, in the case of both emerging and developed countries, limiting the flow of international capital inflows during boom periods would help limit the recession in the event of negative external shocks. [6]In this sense, capital controls could be an effective macroprudential policy in addition to their palliative role in times of crisis.

Conclusion

It should be remembered, however, that even when desirable, capital controls are difficult to implement. On the one hand, because it is not easy to define either the base—the choice of transactions subject to controls—or the desirable rate. On the other hand, because they raise a number of political questions. First, limiting international borrowing in normal times amounts to limiting real wage growth for peripheral countries, which seems difficult to sustain in the short term. In addition, they require the modification of prior agreements in monetary unions, since capital account liberalization is a prerequisite for entry into the euro area. However, unlike competitive devaluations in flexible exchange rates, capital controls are not necessarily policies pursued « to the detriment of neighboring countries, » as Schmitt-Grohé and Uribe point out: in times of crisis, promoting foreign borrowing also means importing more goods produced abroad, thereby supporting activity in the rest of the zone. Farhi and Werning also conclude that capital controls can be considered in an uncoordinated manner, regardless of the heterogeneity of countries within the zone.