Management of Large Capital Inflows: Policy Challenges

Large capital inflows are desirable in principle for relatively low-income countries, because they induce an efficient international allocation of capital and expand the receiving countries' consumption and investment frontiers, allowing simultaneously for more investment and higher consumption levels, and speeding up the growth and real convergence process. However, they also pose potential risks from two sides: overheating and volatility.

The first risk is that of the (in)famous convergence play, a combination of real appreciation and declining long-term interest rates due to falling inflationary expectations and country-risk premiums, which makes the economies even more attractive for short-term capital inflows and portfolio investment. If the demand financed by capital inflows falls entirely on tradables, it can simply be absorbed by large trade deficits. As witnessed by the experiences of Italy, Spain, and Portugal in the late 1980s and early 1990s, convergence play in practice fuels domestic demand for nontradables, too, where domestic supply is limited, leading to severe economic overheating with inflationary pressures. With a fixed exchange rate, the increase in the price level leads to a real appreciation of the domestic currency. With a floating rate, the central bank can do more to suppress inflationary pressures and allow the nominal exchange rate to appreciate.

These conventional demand effects may be augmented by financial market or balance sheet effects (see Calvo 2002, 2003, Calvo et al. 1999, 2004), in what Calvo and Reinhart (1999) call the Fisherian channel of the transmission of capital inflows. The real appreciation of the home currency induces a rise in the relative price of nontradables. The more it rises, the more the central bank tries to stabilize the nominal exchange rate. As a result, producers of non-tradables face a lower ex-post real interest rate and rising cash flows that raise the value of assets that can be collateralized against bank loans. Large capital inflows are, therefore, often connected to asset and real estate price bubbles fuelling credit booms. To the extent that they are absorbed by an expansion of international reserves at the central bank, the ensuing monetary expansion contributes to this development.

We can assess this risk by looking at recent growth rates of narrow money and credit in the new member states (Table 6 [ PDF 40.7KB | 1 page ]).

The table shows the average growth rates of narrow real money and real domestic credit between 1999 and 2003 and between 2004 and 2006. To put them in perspective, the average growth rates of real GDP over the same period are subtracted. Two groups emerge in this table: Bulgaria, the Czech Republic, Poland, Romania, and the Slovak Republic, which had growth rates of real money exceeding real GDP growth by 0-9 percent in the years prior to EU accession, and Estonia, Hungary, Latvia, Lithuania, and Slovenia, where this difference exceeded 10 percent. Following accession, real money growth rose strongly in Bulgaria and Lithuania, but stabilized somewhat in Hungary and Slovenia.

Falling interest rates and declining inflationary expectations may have caused a decline in the equilibrium velocity of money. If the income elasticity of the demand for money exceeds 1, strong real GDP growth adds another explanation. Thus, real money growth rates of 6-8 percent above real output growth may not be excessive. However, the strong monetary expansions in the second group raise a red flag. Turning to credit growth, the ongoing process of financial market development leads to the expectation that credit is growing fast in the new member states. Nevertheless, the table shows four countries with clear signs of strong credit booms, Estonia, Latvia, Lithuania, and Slovenia. Taking money and credit growth rates together, they seem to be the critical cases in the group. This is interesting because, in the past, these four countries also put the largest weight on stabilizing their exchange rates among the countries in this group (von Hagen and Zhou, 2004; Thimann et al. 2004).

To prevent the large capital inflows from excessively fuelling domestic demand, many central banks, especially those with exchange rate pegs, have tried to sterilize them, with the result of a sharp rise in foreign assets held by the central bank; see Figure 11 [ PDF 47.6KB | 1 page ]. This sterilization, however, can be costly, as the foreign assets purchased by the central bank typically have interest rates below domestic-currency denominated assets. The resulting interest payments can become sizeable burdens on the central government budget, and are dubbed “quasi-fiscal costs” of sterilization. For example, central bank losses from sterilizing operations amounted to 2.3 percent of GDP for the Slovak Republic in 2002, and a similar loss was expected for 2003 (OECD, 2004); in Hungary, the central government paid the equivalent of 1.3 percent of GDP to the central bank in 2002 to cover losses from sterilization (MNB, Annual Report 2002). The BIS estimates that the annual costs of sterilizing intervention in individual years between 2000 and 2003 amounted to up to 0.15 percent of GDP for the Czech Republic, 0.2 percent for Poland, and 0.74 percent for Hungary (Mohanty and Turner, 2005). Hauner (2005) points out that the quasi-fiscal costs of holding large amounts of international reserves should include the opportunity foregone to pay down foreign debt or to finance public investment. He estimates the costs of sterilizing capital inflows at 0.5-0.6 percent of GDP annually for the CEE countries. Thus, the costs of managing capital inflows can be substantial.

Figure 11 shows the development of central bank net foreign assets relative to reserve money. For Slovenia, we use currency in circulation multiplied by a factor of 10 to obtain a similar order of magnitude. On the left-hand axis, we show the ratio for the countries that pegged their exchange rates, while for the floating-rate countries (Czech Republic, Poland, Romania, and Slovak Republic), we use the right-hand axis. For the first group, the ratio of net foreign reserves to reserve money was stable or slightly increasing over the 11 years under consideration, indicating that central bank money growth was primarily driven by foreign-exchange policies. Remarkably, for the floatingrate countries, we observe a rising trend in the ratio, indicating that they purchased foreign assets on large scales, presumably to keep their currencies from appreciating faster. Thus, the large capital inflows fuelled monetary expansions even in countries that did not officially peg their exchange rates.

The second risk connected with large capital inflows is their volatility. To date, in fact, capital inflows to the new member states have been quite volatile. Table 8 [ PDF 40.2KB | 1 page ] reports the standard deviation of annual capital inflows relative to GDP between 1994 and 2003. This ratio varied between 2.6 percent of GDP for Poland and 5.0 percent of GDP for Hungary. Volatility is high compared to the average inflows reported in Table 4. The table also shows that several countries in this group experienced large reversals of capital inflows, Sudden stops in the term used by Calvo and Reinhart (1999). Between 1999 and 2000, capital inflows slowed down in seven of the ten countries, the exceptions being the Czech Republic, Estonia, and Slovenia. Between 1994 and 2003, eight of the ten countries experienced at least one year in which capital inflows declined by more than five percent of GDP, while four experienced a decline of (almost) 10 percent or more. This confirms the observation by Calvo and Reinhart (1999) that large capital inflows are often followed by sudden stops and reversals. With the exception of Poland and the Slovak Republic, the reversals reported in Table 8 easily qualify as large compared to the evidence reported by Calvo and Reinhart. Obviously, they have affected countries with very different exchange rate regimes, supporting Calvo's (2003) argument that exchange rate policies are of secondary importance for the incidence of sudden stops. Note also that the largest reversals occurred around the year 2000, confirming the observation in Calvo and Reinhart (1999) and Calvo et al. (2004) that sudden stops are bunched in time and across countries.

Sudden stops create macroeconomic problems through the same channels discussed above in reverse (Calvo and Reinhart, 1999). A sudden stop requires a contraction of the current account deficit or the money supply or both, leading to a contraction in aggregate demand. The ensuing real depreciation of the currency entails a drop in the relative price of non-tradables. Producers of non-tradables now face higher ex-post real interest rates and lower asset values anticipated, including those assets they can use as collateral for borrowing from banks. Banks react to the resulting deterioration in the quality of their loans by cutting back lending. The resulting credit crunch makes the recession more pronounced and longer lasting. In principle, this financial effect can be avoided by a large nominal depreciation of the currency. This, however, increases the burden of foreign currency debt on the government and the private sector.

Coping with large capital inflows is a difficult task for macroeconomic policy. Since the underlying reason is real, there is not much that monetary policy can do. The obvious response is to tighten monetary policy to prevent aggregate demand from overheating. With a fixed exchange rate, capital inflows then lead to a rapid increase in international reserves. The central bank may try to sterilize their impact on the money supply, but in practice this is costly and ultimately of only limited success. Inflationary pressures then result in a real appreciation, a loss in international competitiveness, and a widening current account deficit. Under a flexible exchange rate, the central bank may be more successful in keeping inflation low, but at the cost of a nominal appreciation of the currency, with the same effect on competitiveness and the current account.

At the same time, episodes of large capital inflows into small open economies generate a preference for low exchange rate variability, even if the official exchange rate regime allows for a high degree of flexibility. This has been dubbed the fear of floating in recent literature. The reason is that, since emerging-market countries typically cannot borrow internationally in their own currency, large capital inflows lead to a mounting stock of foreign debt denominated in foreign currency. Exchange rate variations then expose the government and private sector to fluctuations in their balance sheets. Hausmann et al. (2001) show that fear of floating is strongly associated with a country's borrowing in foreign currency and the degree of exchange rate volatility it allows.²² If this is true for the new EU member states, they will show a tendency to tightly manage their exchange rates as the capital inflows continue to persist. Today already, five of the CEE countries have adopted intermediate or hard pegs and Slovenia has joined the Euro area; see Table 9 [ PDF 43.8KB | 1 page ]. They may even decide to enter the ERM-2 for that reason, hoping that it will offer more credibility to their commitment to exchange rate targets.²³ Yet, the comfort offered by an exchange rate peg in this situation can be quite betraying. As the risk of exchange rate variability seems to be low, private borrowers and the government are more inclined to borrow in foreign currencies than they would be otherwise, which increases the exposure to sudden stops and exchange rate crises. As long as the capital inflows continue to be large, the exchange rate peg brings a monetary and credit expansion that aggravates the tendency for overheating. Once the capital flows dry up, the peg may come under speculative attacks, which, unless they can be successfully defended, are costly and more disruptive than the adjustment under a floating rate.

The ERM-2 may offer some relief and credibility in such a situation due to the financial support for interventions it provides, but the history of the early 1990s suggests that its usefulness is limited at best. The experience teaches that European exchange rates tend to become objects of politics, especially in situations where there is market tension. The countries exposed to convergence play failed to adjust their exchange rates in a timely way in the late 1980s and early 1990s, and this contributed to the size of the later devaluations and currency crises. When Germany asked for a revaluation of the DM to absorb the post-unification capital inflows, other governments and central banks were unwilling to grant it. It is not clear a priori that the new member states would not see similar resistance against repeated devaluations of the euro against their currencies, which might be required to counteract inflationary tendencies if capital inflows continue during their ERM-2 membership. Thus, the multilateral nature of the ERM-2 does not obviously add to its economic rationality. It is equally uncertain that the multilateral political negotiations required for devaluations can be completed fast enough in the case of a sudden stop. The multilateral political framework may, in contrast, create ambiguities and rumors in the markets, which can undermine the credibility of the pegs.

Since a sudden stop of capital inflows is equivalent to a cut in international credit to the home economy, the appropriate response by the central bank is to expand credit to the private sector. This can be done through open market operations or loans to the banking system under a flexible exchange rate and entail a nominal depreciation of the currency. The latter also reduces the need for the relative price of non-tradables to fall, but increases the domestic value of the foreign debt burden on the government and the private sector to the extent that foreign debt is denominated in foreign currency. Maintaining an exchange rate peg, in contrast, allows one to avoid the valuation effect, but the loss of international reserves at the central bank leads to a monetary contraction that makes the credit crunch more severe. Thus, sudden stops create a monetary policy dilemma. As recent literature has noted, euro-ization offers a partial way out of this dilemma.²⁴ First, it eliminates the valuation effect on the affected country's debt denominated in euros. Second, the supply of bank credit is no longer limited by the domestic central bank's supply of bank reserves but by the ESCB's supply of bank reserves. This makes any the credit contraction less severe, as monetary policy will not add to it. As a result, countries facing large (and volatile) capital inflows should have a preference for either floating exchange rates or euroization, but avoid soft pegs, especially if, as in the case of the ERM-2, they are unprotected by capital controls.

Fiscal policy is the more appropriate policy instrument for dealing with capital flows. In the face of large inflows, tightening the fiscal stance helps reduce the risk of economic overheating. Here, again, the quality of the fiscal adjustment matters. If tightening is achieved by raising tax rates, the result is buoyant tax revenues and, therefore, a strong temptation to expand fiscal spending. At the same time, initiatives to cut spending in the face of a strong economy will not be very popular. Furthermore, Calvo (2003) points out that, by raising distortionary taxes, the government may reduce the economy's growth potential and thus precipitate a sudden stop. Again, it is important to achieve tightening by cutting government expenditures rather than by raising taxes. This makes the role of good budgeting institutions especially important. Effective spending controls and medium-term fiscal targets that are well anchored in the planning and implementation procedures are important for achieving a sufficient degree of fiscal discipline and using fiscal policy to manage capital inflows.²⁵

As most of the new member states need to tighten their fiscal policies to meet the requirements of EMU, managing capital inflows and meeting these requirements are complementary goals for them. However, the countries with the tightest fiscal stance in recent years are also those that have experienced the strongest credit expansions. For them, as for the others in the future, further tightening to fend off the macroeconomic effects of large capital inflows may be asking too much of fiscal policy (Jonas, 2004).

There is also a task here for prudential supervision and banking regulation involved in managing large capital inflows. Recent empirical studies show that large credit booms and strong real appreciations are among the best indicators of the risk of currency and banking crises.²⁶ Banking regulation can help to prevent capital inflows from spilling over into domestic credit booms (Begg et al., 2003.) Strict rules against overlending and overexposure to individual borrowers are one important element. As lending booms are often triggered by real estate price bubbles, limits on the use of real estate as collateral can serve as another element for protecting the banking system against adverse developments. Furthermore, currency mismatch in the aggregate balance sheet of the banking sector has been an important part of the link between banking problems and currency crises in recent years. Systemic risk arising from large exposure to international interest rate shocks or sudden capital outflows may not be visible in individual bank balance sheets even when it is in the aggregate balance sheet. Monitoring the entire banking sector's financial position is, therefore, an important part of banking supervision in the new member states.

One important feature of the financial opening of the CEE countries in this regard has been the transformation of banking sectors. At the beginning of this process in the early 1990s, the banking sectors consisted largely of a few incumbent institutions emerging from the socialist mono-bank systems in each country. The exception was the Baltic countries. Banking systems were generally quite weak, and they went through a series of crises. In the late 1990s, foreign banks were given permission to move into the CEE countries and quickly established a presence there. In 2006, foreign-owned banks had market shares of at least 50 percent in all the CEE countries, and more than 80 percent in Hungary, the Slovak Republic, Czech Republic, Lithuania and Estonia (EBRD, 2006). This added significantly to the stability of the financial sector of these countries, as the parent banks have provided liquidity and capital support during banking crises, enabling their local subsidiaries to maintain lending when local banks had to cut back their own. Furthermore, the presence of foreign banks has added to the quality of banking supervision, since, under EU rules, supervisors in the parent banks' home countries are responsible for the consolidated institutions. It is, therefore, likely that the presence of foreign banks has strengthened the CEE countries' ability to cope with the macroeconomic challenges posed by large capital inflows.

Download this Discussion Paper [ PDF 228.1KB| 42 pages ].

Post a Comment

We welcome your feedback on this publication. Post a comment. ADBI is not obliged to acknowledge or publish comments and may abridge or edit them before web posting.

Comment(s)

There are [0] comment(s) for this entry. Post a comment.