In transition countries, the issue of how to manage a bank well is intertwined with the issue of rehabilitation. Existing banks are just as likely to be models of how not to run a bank as models of how it should be done.

If I were the CEO of a bank in Croatia today, I might well have some sleepless nights. The CEOs who are getting the least sleep, I would guess, are those at the helm of the banks under rehabilitation, as well as those who may be facing rehabilitation. And while these executives are not sleeping, they are causing unquiet nights for their colleagues in other banks. But the night is young, so let us tell the whole story from the beginning.

Bank failures are common worldwide. We Americans probably boast the largest number of them, since we have the most banks anywhere. Between 1983 and 1993, over 1,000 banks and savings banks failed in the US. In comparison, the handful of Croatian banks under rehabilitation does not seem so extraordinary.

Foreign experiences and models

According to Aristobolo de Juan, the architect of Spain's bank rehabilitation programme in the 1980s and one of the world's leading authorities on bank rehabilitation, there are several mistakes to avoid in dealing with failed banks:

1. counting on inflation to wipe out past losses;
2. merging banks to solve the problem;
3. lasting liquidity support or refinancing of bad loans by the central bank;
4. insufficient recapitalisation that fails to provide an adequate capital base;
5. recapitalisation while keeping the old management in place.

While this list applies to transition countries, including Croatia, transition countries face specific problems. In developed market economies, there are good banks that can serve as models of banking practice. Good borrowers can be found, which is also important; when there are plenty of creditworthy customers around, being a banker is not so difficult.

In transition countries, the issue of building up the skills to run a bank well is intertwined with the issue of rehabilitation. Existing banks are just as likely to be models of how not to run a bank as they are to be models of how to run a bank. And there is the problem of history: who is responsible for what came before? This question has only partly been solved in Croatia; even after the recent London Club agreement, a great deal of old foreign debt, not to mention local debt, remains to be sorted out.

There are several rehabilitation models in vogue among transition countries. Croatia has chosen to follow a modified form of the model used in Spain and Slovenia, in which failed banks are taken over by a government agency, 'cleansed' of bad assets via the replacement of bad loans by bonds, recapitalised and sold. The basic idea of rehabilitation, as opposed to liquidation, is this: failed banks should not be allowed to go bankrupt, whether because they are too big, or because the impact on confidence in the banking system would be too great, or simply because, with deposit insurance, the cost of allowing the bank to fail would simply be too great.

Of course, there are other approaches. Estonia, in particular, has allowed banks to fail, and has refused to guarantee that depositors would get all their money back. Ardo Hansson, a member of the board of the Bank of Estonia, argues that this strategy has greatly improved bank discipline and enhanced the credibility of the Central Bank.

Poland also has adopted a distinctive strategy. Rather than waiting for banks to fail, they devised a way for banks to deal with their bad assets. Banks were required to establish loan workout departments, responsible for handling bad or non-performing loans. When they did this, they gained the right to initiate restructuring or liquidation procedures in debtor enterprises. As a way to enhance bank's ability to force restructuring on unwilling enterprises, banks were given the right to turn loans into shares without the debtor's consent (debt-equity swaps). And to ensure that the banks would come out of the process in good health, the banks were recapitalised. The whole process was carried out under strict time limits, and under the supervision of the National Bank of Poland.

The model adopted by Croatia protects depositors, in contrast to the Estonian model, and puts most of the onus for dealing with problem debtors onto the Bank Rehabilitation Agency (BRA), in contrast to the Polish model.

So far, the Croatian model most closely resembles the model employed in Slovenia. There, bank rehabilitation has been confined to three large banks - Ljubljanska Banka, Kreditna Banka Maribor and Kreditna Banka Nova Gorica. More than three years after the initiation of rehabilitation in Ljubljanska Banka, the following results have been observed. First, interest rates have fallen sharply. In April 1993, just a few months after the start of rehabilitation, Ljubljanska Banka cut its lending rates. Other banks followed suit, and short-term credit fell 4.8 per cent, and long-term credit 7.6 per cent (average real interest rates in the country as a whole). Since then, lending rates have continued to fall, reaching 11.6 per cent real terms, which, it must be said, is still too high.

Second, new management has been installed at each bank. Two of the banks were merged, and all three were 'cleansed' of all assets and liabilities relating to former Yugoslavia. The two resulting banks were given new names and relaunched.

Third, bonds were issued to replace bad assets. Initially, the bonds were denominated in deutschmarks, with a seemingly attractive yield of eight per cent. However, the appreciation of the tolar in 1994, and the fact that the bonds were 30-year and not tradeable, led to protests by the banks in rehabilitation.

They complained of currency and maturity mismatches, as well as inadequate liquidity. This led to the replacement of the bonds with tradeable, shorter-term, domestic currency bonds. These new bonds, maturing in three to 15 years, have yields of 4.5 to 6.5 per cent after revaluation of principal, and seem to meet most of the objections of the banks in rehabilitation.

Fourth, the banks are now making good loans. Both new banks now have a very high percentage of high-quality, 'A' loans on their books, and they both seem to have set aside adequate provisions. Best of all, both were profitable in 1995.

Fifth, the bank programme was supported by a loan from the World Bank, and advisors from such institutions as the US Resolution Trust Corporation and Allied Irish Bank provided help.

Got rich overnight

The rehabilitation process, initiated in November 1995 at Slavonska Banka (voluntary rehabilitation) and early 1996 at Rijecka Banka and Splitska Banka (involuntary rehabilitation), has already had a drastic effect on the money market. With these banks no longer desperate for daily liquidity, demand on the money market has shrivelled, and interest rates have dropped from nearly 32 per cent to nearly eight per cent.

This is certainly good news for Croatian bankers needing to cover minuses on their daily accounts. However, there is a bad side too: many banks had made significant profits easily on the money market. Those days are over. And this is not just an insignificant sideline for banks: overnight loans accounted for some 54 per cent of total bank credits in mid 1996. Considering the high rate achieved on overnight credits, and the fact that these credits, unlike many others, were actually regularly repaid, the blow to bank profitability could be substantial. Furthermore, banks will have to work harder to make loans, and will incur greater risk in doing this.

As the restructuring and rehabilitation process proceeds, the rehabilitated banks will start to be more influential in other markets too. Since these banks are all heavyweights, it can be expected that they will raise the level of competition substantially. They should be in a position to compete for business, presumably by lowering lending rates, as Ljubljanska Banka did in Slovenia. This would be a main benefit of rehabilitation: reasonable interest rates that allow entrepreneurs to fund sound projects, not just the few get-rich-quick schemes that flourish in a situation when interest rates are 25 to 30 per cent or more.

Of course, all this presupposes that de Juan's rules are followed. If the rehabilitated banks don't get enough capital, or if too few of the bad assets are recognised and removed, these banks will continue to be invalids. Which their competitors might not mind, but taxpayers and depositors will not appreciate, since this would just mean future - and very possibly more expensive - bailouts.

Furthermore, the Croatian practice of 'voluntary' rehabilitation could be a dangerous thing. 'Voluntary' rehabilitation allows bank management to 'voluntarily' accept public money, and remain in management of the bank. But is current management really not at all responsible for the situation in the bank? And is current management really able to lead the bank in a totally new direction, to make wrenching changes among personnel, in the company's culture and strategy? If the same management fails again, guess who pays the bill?

Who cares for competition?

Maybe the managers of banks undergoing voluntary rehabilitation will sleep well. But maybe even they will be up when the stars are twinkling. For foreign banks will not stay out of Croatia indefinitely. In fact, they are currently showing their interest. If foreign banks enter in force, they could have a strong impact, as the case of Hungary shows. There, foreign banks have won over some of the best corporate clients, leaving local banks in a difficult position. In other countries, like the Czech Republic, the entry of foreign banks has been limited, for just this reason.

There may come a time when foreign entry needs to be limited, at least to some extent. But right now, the entry of foreign banks, whether on their own or as partners of domestic banks, probably should be encouraged. The benefit done by their ability to provide expertise and to heighten competition will probably outweigh any harm they do in 'stealing' good customers.

Banks in Croatia sorely need competition to reach European standards. Croatian banks average about DM900,000 assets per worker, a rough measure of efficiency. Slovene banks average DM1.2 million, Czech banks DM1.7 million, and US banks DM3 million. Croatia has a long way to go even to reach the standards of the leading transition countries.

And that is not all. In 1997 new bankruptcy laws will come into effect. It would not be surprising if many seemingly profitable and well capitalised banks were to find themselves in difficulty when their clients go bankrupt, and they can no longer pretend that everything is in order with the credits they granted or the equities they hold. As the shakeout begins in the real sector of the economy - and with 7.5 billion Croatian kuna (HRK) of unpaid bills, some sort of shakeout is inevitable - the banking sector will not be unscathed.

Bank crises, according to Estonia's Ardo Hansson, are inevitable in transition countries. Perhaps so. Still, I do not want to cry wolf. Many things are going in the right direction - the beginning of rehabilitation, strengthening of prudential regulation, decreasing interest rates in several segments of the financial markets. And there is certainly reason to think that peace in the region, and normalisation of relations, can bring an economic upturn, which should make bankers' lives easier.

Nonetheless, if I were a Croatian banker, I might well be gazing into the night sky a little more than usual.

The author is a Professor at an American university, a frequent visitor to Croatia over the last ten years, and now a visiting scholar at the National Bank of Croatia

This article first appeared in the Croatian financial magazine 'Banka'