Thanks to having the lowest inflation among transition countries, strong economic recovery, a stable exchange rate, low indebtedness and an ability to control the external deficit (BOP current account deficit decreased from 6.8% of GDP to 4.5% in 1996), in early 1997 the Republic of Croatia won BBB- or Baa3 sovereign ratings. External debt was USD 4.8 billion (24.3% of GDP) by the end of 1996, and the central bank’s international reserves reached USD 2.3 billion (2.8 average monthly imports of goods and services). With an addition of USD 2 billion of commercial banks’ liquid foreign assets, Croatia’s net foreign debt was USD 502 million i.e. 2.5% of GDP ¹ . (see the appendix).

Twelve months later, external performance deteriorated sharply. The current account deficit to GDP ratio increased more than seven percentage points, and reached 11.8% for 1997. The stock of external debt rose by USD 1.9 billion and reached 34.3% of GDP. Nevertheless, capital inflows were so strong that the central bank’s international reserves increased to USD 2.5 billion (2.7 times monthly average of imports of goods and services) by the end of 1997. With an addition of USD 2.3 billion of liquid foreign assets held by commercial banks, net foreign debt reached USD 1.8 billion i.e. 9.3% of GDP. Croatian foreign indebtedness is still moderate. The exchange rate remained stable, inflation is low, and growth continued into early 1998. However, the size of the current account deficit, as well as the magnitude of its increase during 1997, raised the problem of vulnerability. Another important related question is: how much did good credit ratings contribute to the possibility to increase net foreign indebtedness by 10% of GDP in only one year?

Events in Croatia in 1997 showed that ratings matter, especially when doors are open to capital inflows. Economics is not a repeatable science, so one can not go one year back in time and repeat the Croatian experiment without the ratings in order to see how much of net loan inflows would have occured without an investment grade. But if one assumes at least some price elasticity of Croatian demand for foreign funds, it is reasonable to expect slower accumulation of foreign debt. In general, autonomous economic forces that work only via the impact on the current account (terms of trade, restructuring in the tradable sector, productivity gains) became the effects of the second order. Impacts of the first order came from the financial account, which is more volatile, harder to predict and subject to sudden changes in moods and perception. The importance of these factors is widely recognized regarding their impact on fluctuation of the short-term foreign investments, but Croatian experience shows that this can also be the case with loan inflows of a more favorable term structure.

Predictions /learning about economy by analogies

Current account to GDP ratio

Structure of foreign financing

Term Structure of Foreign Debt

Are Credit Ratings Part of Economic Fundamentals?

Credibility of Economic Policy: Trust and Banking

Conclusion

Appendix: Main Indicators

Credit ratings are very important for a country like Croatia. It helps to overcome the lack of private business information about the country. However, winning or loosing the grade may imply large financial fluctuations if a country is open to capital inflows.

Investors’ sentiments about emerging markets became more volatile after the Asian crises. Every country which is in position similar to Croatia, has to be aware of financial markets’ perceptions of information about the country. Perception formation has two stages: learning about countries and making predictions about them. Markets usually do it by analogies regarding cause and effect: if a certain indicator or event(s) occurs prior to an unfavorable economic event (debt crisis and/or currency crisis and/or recession) in one country, markets tend to punish other countries when similar events occur or when certain indicators have similar values as in a country which experienced problems. Investors punish (liquidate investments in) the country even if it is not clear why the chain of cause and effects has to be the same in different countries. Here I will elaborate about four crucial indicators: (a) current account balance to GDP ratio, (b) structure of financial flows, (c) term structure of foreign debt, (d) quality of the banking sector (elaborated in the section about credibility and monetary policy).

Whenever the deficit to GDP ratio raises significantly, it "rings a bell" in market analysts’ offices. Almost every analyst has learned a threshold of tolerance for that ratio. Of course, there is a lot more information used in the analysis before the conclusion is reached. But the fundamental problem with the deficit to GDP ratio is that it depends on history, size and openness of the country, which sets a strong limit to its uses. Here is a simple example:

Current account deterioration from a 4.5% of GDP deficit to 11.8% occured in Croatia in 1997 because current account credits (US dollar equivalent) increased by 2.1% and current account debits by 14.4% in comparison to 1996. ^{2 3} In 1997, current account credits made up about 50% of GDP and current account debits about 62%, so that the openness was 112% of GDP. If Croatia was larger and twice less open, so that the two ratios were twice lower in 1997 (say, credits equal to 22% of GDP and debits to 34% of GDP), the same rates of change of credits and debits in 1997 (2.1% and 14.4% respectively) would imply only around a 3.5 percentage points deterioration instead of the seven percentage points which actually occurred. Since the current account deficit to GDP ratio fluctuates more intensively as a country is more open, any policy action aimed towards external performance or any external exogenous shock will lead to larger fluctuations of the deficit to GDP ratio in small and open economy, compared to larger economies. This fact is neglected sometimes.

After the Mexican crises in 1994/95 more attention has been devoted to the financial account of the balance of payments i.e. the quality of financing for the deficit. In 1996, the Croatian current account deficit of 4.5% of GDP was financed primarily with FDI. The FDI to current account deficit ratio was 60%. Portfolio investments were negligible, and net loan inflows were 46% of the current account deficit, contributing largely to a significant increase in international reserves. ⁴ But that was the time when Croatia did not have credit rating.

In 1997, the FDI to deficit ratio fell to 16%. Net FDI fell even more strongly, from 58% of the deficit in 1996 to 8% in 1997. Again, here is a problem of a small and open economy, where individual transactions can make a difference at the aggregate level. The Croatian pharmaceutical company Pliva, which is listed on the London Stock Exchange, made a large investment in a plant in Poland. ⁵

Portfolio investments increased from negligible amounts in 1996 to USD 174 million in 1997. This is still a relatively low amount ⁶ , because the range of attractive liquid instruments is limited to short-term government and central bank papers, and shares of Pliva and Zagrebacka banka, listed both in Zagreb and London. The significant rise in portfolio investment largely compensated for the drop in FDI in 1997. So, overall foreign (direct and portfolio) investment to Croatia amounted to USD 521 million (22.8% of the current account deficit) in 1997., which is only 6.5% less inflows (in US dollar terms) than in 1996. ⁷

The new credit ratings and the fundamentals lying behind it, undoubtedly helped to promote portfolio inflows in 1997. But, the impact of ratings on loan inflows was much stronger. In 1996 net loan inflows were USD 404 million. In 1997 they increased by 457% in comparison to 1996, and reached USD 2251 million i.e. 11.6% of GDP. ⁸ 26.5% of net loan inflows was directed to the government sector and it was directly linked with sovereign ratings granted at the beginning of the year. 28.7% of net loan inflows was directed to Croatian commercial banks. Two Croatian banks won international credit ratings in 1997. The rest of the 44.8% of total net loan inflows was directed towards the rest of the private sector, mainly enterprises. Some large companies were able to raise foreign debt without government guarantee (the largest oil company, electrical public utility, etc.) and some won corporate ratings.

In conclusion, credit ratings promoted huge net inflows from abroad in 1997. Most of it occurred as net loan inflows. Speculative capital inflows was not significant, and its fluctuations can be accomodated by changes in a small fraction of international reserves (see footnote 6).

Following up the story about learning and making predictions about countries by analogies, what next comes to mind is the term structure of the debt. Out of USD 2251 million of net loan inflows in 1997, 470 million i.e. 21% was short-term debt. The stock of short-term foreign debt remained bellow 10% of the total stock of foreign debt. Even very strict definitions of short-term foreign "debt" (including the stock of all accumulated foreign portfolio investments) show that short-term foreign liabilities will remain below 25% of the total liquid foreign assets of the banking system in 1998.

In conclusion, term structure of Croatian foreign debt shows low exposure to sudden changes in markets’ sentiments.

I believe that sovereign ratings should be understood as an endogenous part of economic fundamentals, at least for emerging market economies. Winning an investment grade for the first time in an emerging market which had just begun integration with world markets, naturally leads to greater financing opportunities and a larger current account deficit. In a (very) small and (very) open economy which begins financial integration, this situation can lead to large fluctuations in the current account deficit to GDP ratio, as shown by a simple numerical example above.

However, my feeling is that investors mostly don’t see the ratings process as endogenous. They tend to underestimate its impact on the economy. As they tend to learn and make predictions about countries on the basis of cross-country analogies, the mere fact that the current account worsened significantly for some investors is enough to raise doubts and liquidate investments. If these types of investors prevail, their expectations can become self-fulfilling prophecies. New economic reality can be created unintentionally.

Local policy makers have to be aware of the problem. They have to find ways to communicate messages to markets and investors. They have to keep in mind that markets: (a) learn and make predictions mainly by cross-country analogies, (b) do not view the credit ratings process as an endogenous part of fundamentals, but rather look at it as an objective judgment. There is no problem when policy makers’ and markets’ opinions coincide, like in this case: the deficit of 11.8% is too high and should be reduced significantly. But what if by the end of 1996 policy makers in Croatia announced that they wanted to increase then existing deficit of 4.5% of GDP even higher, which was not a bad policy, given very low foreign debt at the time? Was there any possibility to communicate this message to the markets and have it accepted by the markets? The answer depends on credibility of economic policy.

Probably both local policy makers and market participants are more or less aware of this need to communicate information and built mutual trust and understanding. The problem for emerging market economies is that there are no institutions with a sufficiently long history and reputation to build trust. Essentially, this is the credibility problem.

Credible economic policy is one which recognizes all fundamental problems and acts upon all of them in a mutually consistent manner. Economic literature emphasizes mutually consistent but independent monetary and fiscal policy. The literature about stabilization adds incomes and foreign exchange policy to the overall policy mix which has to be consistent. Restructuring efforts in both corporate and banking sectors is added to this list in developing countries. Credibility is needed in order to make public expectations consistent with policy measures. It is a sufficient condition for the success of economic policy.

The credibility problem is at the very core of the current economic problems in Croatia. Namely, by mid-1997, policy makers in Croatia learned that the current account deficit was widening. They clearly perceived the growth of both wages and domestic credits of the banking system going above the targets. Monetary policy became restrictive as of August 1997 in order to curtail credit growth (see figures 1 and 2). This policy did not work because it did not influence all sources of financing of credit growth. At the same time, banks continued to finance credit expansion by decreasing their foreign assets and/or increasing foreign liabilities (borrowing from abroad). Enterprises (state-owned enterprises especially) continued to borrow from abroad too. Both banks and enterprises rushed to achieve ambitious growth targets despite of changes in the monetary policy stance which is obvious from the dynamics of the money multiplier (figure 1). The result of this ambition was that 60% of total net loan inflows for 1997 occurred in the last quarter, when monetary policy was already tightened. Moreover, 95% of total net loan inflows of the banking sector occurred in the last quarter.

Banks' behavior regarding credit policy did not change significantly in the first quarter of 1998 (see figure 3), so the central bank decided to impose new capital restrictions in April 1998. Obligatory kuna deposits have to be held by banks with the central bank in an amount of:

a) 10% of foreign banks' interbank deposits with domestic commercial banks,

b) 10% of the value of guarantees issued by banks for foreign borrowing of their clients up to three years,

c) 30% of the value of short-term borrowing abroad by banks.

According to data for April and May, these measures were very effective. The trend of falling net foreign assets has stopped (figure 3). Rates of growth of total credits to the private sector have been falling rapidly (figure 4) and now they have been brought in line with policy makers' targets. In addition to these, international trade statistics data for the first four months of 1998 showed that the trade deficit is falling. In US dollar terms, the trade deficit for January to April 1997 was 1176 million, and for the same period 1998 1140 million i.e. 3% lower (figure 5). ⁹ As inflation remained under control with good prospects to remain the lowest among transition economies (figure 6), it appears that the credibility of macroeconomic policy has been restored in a very short period of time.

However, two processes render the broad picture still filled with uncertainity. The first is the exchange rate process. As banks and enterprises tried to achieve their growth targets despite of a change in the monetary policy stance (by taking loans from abroad in the last quarter of 1997), they induced unexpected exchange rate appreciation (figure 7). The kuna traditionally gets stronger during the summer tourist season and then remains flat or even depreciates slightly. However, foreign loans inflows were strong enough to induce appreciation in the last quarter of 1997. When the net inflows slowed down and stopped under the influence of new capital restrictions, the currency started to depreciate. The depreciation versus D-mark was still below 1% in March, but in April and May it reached 1% monthly.

The second process which burdens the Croatian economy at the moment is related to banking sector problems. The fifth Croatian bank by size, privately-owned Dubrovacka banka, failed in March. The government decided that it was big enough (and important enough for the local economy in the south of Croatia) to be rehabilitated. ¹² Owners lost their stakes (losses exceeded the liable capital of the bank), and management changed in April. Dubrovacka banka was a textbook case of failure: rapid growth, connected lending, lack of internal controls that raised doubts about possible criminal acts (the former president of the bank was imprisoned) and speculation about political connections.

The Dubrovacka banka case raised doubts about the stability and solvency of other banks (especially medium-sized) that look similar to Dubrovacka. The broad public suddenly became very sensitive to anecdotal information about connected lending that appeared in the press. Consequently, two other medium-sized banks experienced moderate deposit withdrawals. Up to now, they have accomodated successfully because they held some free reserves and still have credibility at the interbank money market. However, other banks correctly perceived higher risks, and the money market interest rate began to rise sharply from March onwards (presently, it is around 17%).

A lack of public information about individual banks at times when public sensitivity has increased sharply, makes the credibility problem very important. Policy makers are trying to rebuild confidence by stating that the problems are on the "edge" of the banking system. The "core" of the system is healthy: four large banks underwent rehabiliation procedures, one of them has already been privatized, the largest bank (Zagrebacka banka, privately owned, listed both in Zagreb and London) published very good results for 1997, and a number of medium-sized banks which are traditionally ranked well, are doing fine. There are problem banks, but they cannot threaten the system. Their problems will be resolved within a few months, including a high probability that some of them will exit the market.

These messages induced different investors’ reactions. Depositors did not withdraw money from their accounts on a large scale. Actually, they were shifting deposits from banks that are (more or less correctly) perceived as problem ones to other banks. In a sense, this is a healthy economic process. However, banking sector problems (coupled with exchange rate depreciation) stopped the inflows of new foreign exchange savings (repatriation of foreign exchange held by residents at accounts abroad). This can be interpreted as a sign of unstable confidence which can be eroded further if banking sector problems won’t be resolved speedely and in a transparent way.

Equity investors (domestic and foreign) expectedly reacted in a more furious way. The Zagreb Stock Exchange Index (CROBEX) has fallen steadily since the beginning of 1998, banks’ shares being the greatest loosers. Equity markets have been punished on the basis of general information about banking sector problems and external deficit in 1997.

There is significant room (and need) for policy makers to rebuild confidence and credibility by taking immediate actions in the banking sector. The macroeconomic setup is favourable for building up credibility: net foreign assets of the banking system stopped falling, credit growth is slowing down, preliminary trade figures have begun to show lower deficit in comparison to the same period last year, and depreciation will likely stop in the very near future. The final step which will coordinate information and better expectations is needed on the banking front: markets want to know which players are good and which are bad; they need to know what will happen to bad ones and when.

Credit ratings should be viewed as part of economic fundamentals for an emerging market economy which started to integrate with world financial markets by liberalization of current and capital account transactions. Croatia won an investment grade in early 1997. It helped to promote foreign portfolio investment and, more importantly - by significant loan inflows. Croatian experience shows that an emerging market economy can (at least could, before the Asian crisis) attract huge amounts of loan and non-loan inflows in a very short period of time.

Strong inflows induced a widening of the current account deficit which was very pronounced because Croatia is a small and open economy. A large deficit threatened confidence, because markets do not perceive a widening deficit (at least partly) as a consequence of an investment grade given to a small, open economy, which just began financial liberalisation. A large and increasing deficit is viewed by markets rather as a sign of more fundamental weaknesses. So, despite high international reserves and very low net foreign debt (10% of GDP), policy makers in Croatia decided to cut the deficit strongly in 1998 in order to rebuild confidence. Restrictive monetary policy was combined with newly implemented restrictions to capital flows, and this policy proved to be effective as credit growth began to slow down significantly in April and May. The last step in the process of new confidence building is consolidation of the banking sector, as some of the rapidly growing aggressive banks suffer from connected lending and other fundamental weaknesses. As macroeconomic policies are set up properly now (trade deficit beginning to improve), and the core of the largest banks are healthy, it is likely that credibility will soon be rebuilt.

 

	1994	1995	1996	1997	1998f*
Real GDP (%)	5.9	6.8	6.0	6.5	6.0
Inflation (%)**	-3.0	3.7	3.4	3.8	5.0
Fiscal balan./GDP (%)	1.7	-0.9	-0.5	-1.4	-0.4
Exports f.o.b.***	4260.4	4632.7	4545.8	4376.4	4726.5
Imports f.o.b.***	5582.9	7870.2	8197.0	9448.9	9165.4
Trade deficit	-1332.5	-3237.5	-3651.2	-5072.5	-4438.9
Balance of services	1661.9	1207.7	1764.0	2022.1	2386.1
Current account balance	785.9	-1283.2	-880.8	-2282.9	-1496.0
Current account/GDP(%)	5.4	-6.8	-4.5	-11.8	-7.0
International reserves of the CNB (USD million)	1405.0	1895.2	2314.0	2539.0	2700.0
Liquid foreign assets of commercial banks	902.4	1369.5	1992.1	2333.2	2400.0
External debt	2821.5	3336.4	4808.4	6661.6	7600.0
External debt/GDP(%)	19.3	17.7	24.3	34.3	35.3
Net external debt/GDP (%)****	3.5	0.4	2.5	9.3	11.6
ST foreign debt/total foreign debt (%)*****	1.8	6.8	8.9	8.1	8.9
ST foreign debt / total reserves (%)******	2.6	7.3	10.5	15.5	22.3
FDI /GDP (%)	0.7	0.5	2.7	1.8	2.2

 

*	Author's projections for 1998.
**	Increase in the rate of inflation is due to one-time effect of VAT introduction, estimated to be around 1.5%. Fundamental inflation remains around 3.5%.
***	Commodity exports and imports in million USD, as recorded in the balance of payments.
****	Foreign debt outstanding minus sum of international reserves of the central bank and liquid short-term foreign assets of commercial banks.
*****	Includes only short-term foreign liabilities as recorded in the foreign debt data base.
******	Short-term debt is increased for cummulative foreign portfolio investment.