der systemischen Bankenkrise nach dem Modell Norwegens (1988 und später)
Norges Bank Conference on Banking Crisis Resolution - Theory and Policy, Oslo 16 June 2005
The Norwegian Banking Crisis affected almost 2/3 of the banking system and led to the nationalization of our three largest banks. At the time, it was the first systemic crisis in an industrialized country since the 1930s. This also explains why the crisis caught most policymakers off-guard.
How could this happen in a well organized, mature economy? How was the crisis resolved? And, what can we learn from the crisis resolution methods adopted in Norway? These are the questions I will address today.
Some may argue that it is difficult to derive a "crisis resolution blueprint" from a banking crisis that was so much influenced by the deregulation of that time. The next crises, if any, will no doubt have other causes. But the resolution methods we adopted may still be relevant for others.
In order to understand how the Norwegian banking crisis could take on systemic proportions, let me first fill you in on the background and present some of the special features of the crisis. Then I will review the resolution methods used, and finally discuss the relevance of the Norwegian experience for crisis resolution today.
Special features of the Norwegian banking crisis
The Norwegian banking crisis was a classic boom-bust crisis with some special, national features:
* Deregulation and liberalization paved the way for the boom
* Macroeconomic policies were largely pro-cyclical
* Lending growth became exceptionally strong
* Prudential capital regulations were relaxed [and]
* Supervision efforts were reduced
This may sound like "a deadly cocktail" [which in fact it turned out to be], so let me comment briefly on each feature:
The banking crisis had its roots back in the late 1970s. Tight regulations of credit and interest rates were in place. High inflation and favorable tax treatment of nominal interest expenses led to highly negative real interest rates in the early 80s. Demand for credit increased steeply, and credit regulations became less effective due to a mushrooming grey market for credit and loans from abroad. This - together with the international trend of liberalization at the time - motivated the deregulation process. So, in early 1984 - the government started a process of swift removal of credit regulations.
Monetary and fiscal policies
Macroeconomic policies became largely pro-cyclical during the 80s. The economy started to grow very strongly in 1983 - following the international downturn in the early 80s. Private consumption grew particularly fast during the mid-80s, with real growth peaking at ten percent in 1985! The government was not aware of this at the time; and estimates of consumption growth were much lower. Despite strong growth, there was also political pressure to keep interest rates low, and attempts to tighten fiscal policy were unsuccessful.
The removal of the credit regulations - while keeping interest rates at a low level - triggered an exceptionally strong growth in bank lending. Collateral was not a problem, as prices soared in the newly deregulated secondary market for housing. The new loans were to a great extent provided by the three largest banks. Their balance sheets increased by 150 percent in just three years.
Prudential capital regulation
The sharp increase in lending was accompanied by an erosion of banks' capital base. The quality of capital was eroded, as the authorities allowed the banks to issue large amounts of perpetual subordinated debt; debt that was accepted as ordinary equity, but in fact was of little use when the crisis hit [as it could only be called upon if the banks had been closed]. The capital buffer was weakened further by inadequate provisioning.
While banks were having a lending bonanza, on-site inspections decreased sharply - from 57 in 1980 to 8 in 1987! There had long been plans to establish a single supervisory agency, and the new agency was formally established in March, 1986 and became operational in 1987. As a result, supervisors' attention was focused on building the new agency, rather than checking on the banks' risk profiles during the on-going lending boom.
The banking crisis
With the benefit of hindsight, the banking crisis was an "accident waiting to happen". And sure enough, when the economy was hit by a strong negative shock and a cyclical downturn, loan losses and non-performing loans soared, wiping out the capital of many banks.
The negative shock came in the form of a sharp fall in oil prices in the beginning of 1986. This led to a run on the Norwegian krone and a devaluation of the currency. Fiscal policy was then tightened sharply and interest rates were raised to restore credibility. Subsequently, the private sector consolidated its financial position through debt repayments, resulting in the deepest recession in Norway since World War II.
The sharp correction in the real economy led to large loan losses in the banking sector. Some banks experienced solvency problems already in 1987 and the problems increased as the recession deepened. Several smaller banks were merged with larger, still solvent banks in this period. The banks' guarantee funds issued guarantees or infused capital to facilitate these mergers. Thus, until late 1990, the crisis resolution was mainly financed by the banking sector and the banks' guarantee funds.
The resolution of the Norwegian banking crisis
By the end of 1990, key policy makers started to worry about the crisis becoming systemic. It was essential to maintain confidence in the financial sector and avoid a major credit crunch when the economy was already in a recession. The capital in the banks' guarantee funds was about to run out and it was apparent that the larger banks might also face problems. The Ministry of Finance thus proposed - in January 1991 - the establishment of a new guarantee fund - the Government Bank Insurance Fund and put 5 billion kroner in it [equivalent to around 0.6% of GDP].
The authorities were at that time actually ahead of events. The banks did not seem to be aware of what was under way. Thus, early in 1990, senior management of the largest bank - Den norske Bank - still expected a positive result for the year as a whole, and the chief executive officer of the second largest bank - Christiania Bank - was voted the best leader of the year.
The new fund was set up as an independent legal entity, located in and supported by technical staff from the central bank. Its mandate was initially to provide loans to the banks' guarantee funds to enable them to perform their roles. During that first year, another eight small to medium-sized banks received support.
Certain amendments were also made to the banking laws, which made it possible for the government to write down a bank's shares to zero. This ensured that share capital really could be written down if a bank's capital was lost.
At this stage, the crisis escalated rapidly. In October, it became evident that huge loan losses had wiped out the entire share capital of two of the three largest banks [Christiania Bank and Fokus Bank]. Their share capital was written down to zero by government decision. The government - through the Government Bank Insurance Fund - became the sole owner of the two banks.
By the end of 1991, most of the share capital in the largest commercial bank [Den norske Bank] was also gone. The bank received support from the government in the form of preference shares in early 1992. This was not enough, and in connection with further capital support later that year, it was decided that the bank's ordinary shares would be written down to zero. When the Government Bank Insurance Fund provided additional support to the bank, it became the sole owner.
Key features of the Norwegian crisis resolution
There are five features of the Norwegian resolution I would like to highlight:
* Private solutions were explored before the government intervened.
* Share capital was written down to zero before committing public funds.
* The government acted swiftly to limit contagion, but did not provide a blanket guarantee.
* Liquidity support was given to illiquid, but solvent institutions.
* The government did not use an asset management company - as the other Nordic countries did later on.
I shall briefly comment on each of them:
Private solutions to problem banks - with financial support from the banks' guarantee funds - were the norm before the crisis became systemic. The funds were - as already noted - quite active in the early phases of the crisis. Membership was [and still is] mandatory and the funds were financed ex ante by fees from the member banks.
The banks' guarantee funds had wide mandates to explore least-cost solutions, like recapitalizing a bank or providing guarantees or financial support to facilitate a merger or a take-over. As you have heard, by the end of 1990, the banks' guarantee funds were unable to handle the crisis as the larger banks started to face problems.
Before the Government Bank Insurance Fund injected capital into crisis-stricken banks, efforts were made to attract private investors. However, these efforts did not succeed. The option of foreign takeovers was also considered by the authorities, but at the time this was not a viable option, due to current banking regulation and strong political preferences for national ownership of banks.
Banks' share capital was written down to zero
Before the government committed any funds to the banks, it made sure that capital was written down according to the best estimate of loan losses at the time. Strict conditions were also tied to all capital support from the Government Bank Insurance Fund. Thus, public support did not come as "free lunches" for the banks, and their activities were curbed to avoid giving them an unfair competitive advantage. Shareholders were forced to cover the bank's losses before tax payers' money was put on the table. To my mind, this was imperative in order to muster the necessary political support and acceptance for the rescue operations.
It should be noted here, that the "zeroing" of equity capital in the three largest banks was hotly debated at the time. So much so, that the Parliament established a commission after the banking crisis was over to address the issue. Their overall conclusion was that the public authorities handled the financial crisis quite well, although the writing-down of the share capital in Den norske Bank could have been handled better. However, they also noted that there was no private domestic capital forthcoming at the time, and no political support for foreign ownership of the bank.
No blanket guarantees were given
Once the crisis had become systemic, the Norwegian authorities acted swiftly to limit contagion. But they did not issue any explicit blanket guarantee for the banks' liabilities. However, in late 1991, the Ministry of Finance announced that the government would implement measures necessary to secure confidence in the Norwegian banking system, while Norges Bank announced that it would secure the necessary supply of liquidity to the banking system. No assurances were given that individual banks or bank creditors would be rescued. In the end, only two small banks were closed and all depositors were repaid in full through the banks' guarantee funds. Creditors also recovered their funds in full, except for some creditors in one of the small commercial banks that were closed.
It is worth mentioning that ex ante, there was not a full deposit guarantee during the banking crisis, as the commercial banks' deposit scheme could in principle reject support measures. However, ex post, all depositors were repaid in full. After the banking crisis, a new deposit insurance law was enacted with limited [although generous] protection.
Liquidity support was only given to illiquid, but solvent financial institutions
Norges Bank was a fairly active lender of last resort during the banking crisis. By and large, it stuck to the classic principle of only lending to illiquid, but solvent institutions.
In the first stage of the crisis, the bank extended liquidity support to several small and medium-sized banks. These loans were guaranteed by the banks' guarantee funds. Most loans were provided at market rates, although a few were provided at below market rates.
In the second stage of the crisis, once it had become systemic, a division of responsibility between the different authorities was established. Norges Bank contributed loans to institutions that were experiencing liquidity problems, but where underlying solvency was satisfactory, while the government insurance fund provided solvency support.
It is also worth mentioning, that Norges Bank entered the banking crisis period with large, unsecured claims on the banking sector after sterilizing the heavy interventions during the currency crisis in 1986 in order to support the government's "low interest rate policy". At the end of 1986, central bank lending to banks thus stood at 14 percent of the banking sector's total assets. These loans were not repaid until 1993, i.e. after the end of the banking crisis.
No use of asset management companies
The authorities chose not to set up separate asset management companies - or "bad banks" - to handle problem loans. Bad loans were not considered to be excessive for management to handle "within" the bank. A bank knows its own borrowers best and has experience in working out defaulted loans. In addition, setting up an asset management company and transferring bad loans from the banks would have required complicated accounting and legal work, and it would have been hard to find a fair price at which the loans should be transferred. Last, but not least, such an asset management company would have required government funding. Thus, more taxpayers' money would have been put at risk. However, there was nothing to prevent the banks themselves, alone or jointly, from establishing subsidiaries to handle the bad loans. None chose to do so.
Norway avoided a collapse of the banking system and an ensuing major credit crunch. The government recognized the severity of the problem at an early stage and showed a willingness to take the necessary measures. As a result, confidence in the financial system was quickly restored. Collateral values started to increase and loan losses fell. By late1993 the crisis was over, the banking sector was again profitable and started to raise private equity in the market.
Due to the quick resolution, the Government was able to contain the fiscal and economic costs of the banking crisis. The fiscal resolution costs have been estimated to about 2 % of GDP, which is low compared with many other crisis countries. However, the net fiscal cost turned out to be negative - the Government actually made a small profit of 0.4 % of GDP - when the government's shares in the banking sector were sold at a substantial profit.
The government became "owner of last resort" at the peak of the crisis when private investors were unwilling to invest. At that time, no one knew where the economy was heading. It was therefore reasonable that the government should benefit from its investment as the economy recovered.
We have also tried to estimate the economic costs of the banking crisis. As you all know, these are difficult to identify and measure, as banking crises often appear at the same time as a recession. Separating the effects of a banking crisis from the general business cycle is a major methodological challenge. Skipping all the qualifications, let me just note that our best estimate of the output losses associated with the banking crisis in Norway is around 10 percent of GDP.
Comparison with the other Nordic countries
In much of the literature on financial crises, the Nordic banking crises are regarded as one crisis, or three rather identical ones. However, the three crises differed, even if they had many common features. The banking crisis started much earlier in Norway, but the crises in Sweden and Finland became much deeper. It also took longer for the banking sectors to recover in these two countries.
The resolution methods were also quite different:
* The bank guarantee funds were actively involved in the early resolution efforts in Norway
* The Norwegian authorities did not issue a blanket guarantee
* Asset management companies were not used in the Norwegian resolution [and]
* Shareholders' equity was written down to zero before public funds were committed to a larger extent in Norway than in the two other countries.
It is instructive to compare loan losses in these three countries with the Nordic country that avoided a crisis - Denmark. Loan losses were in fact as high in Denmark as in Norway, but they were somewhat more spread-out in time. Denmark thus avoided a banking crisis because the banks' capital base was stronger allowing them to absorb losses more easily. This was in part due to favorable tax treatment for general provisions and stricter capital requirements applied by the Danish authorities. This allowed the banks to weather the storm, while Norwegian banks collapsed as losses mounted.
Relevance of the Norwegian resolution policy today
Can we learn something from the crisis resolution methods adopted in Norway? Well, I think so - even though the Norwegian crisis had many idiosyncratic features.
Let me just focus on three important issues:
* The role of shareholders
* The role of the government
* The role of the central bank
I will address them briefly in turn:
Role of shareholders
In a financial crisis, the owners - together with management - are expected to handle the problems. This is obvious to all of us now, but this has not always been the case. Due to tight ownership regulations, banks became institutions without strong, responsible owners. Management tended to run the bank - and not always for the better, as the banking crisis showed. Hence, when the crisis broke, no clear direction was provided from the owners or the general assembly. The government had to intervene in order to save the banks as going concerns.
This has changed today. Ownership regulations have been relaxed. Today owners will normally get permission for stakes up to 25 percent. Foreign-owned subsidiary banks are in fact 100 percent owned. I look favorably on this development, as greater concentration should lead to greater responsibility - especially when the going gets rough.
During the crisis, banks' equity was written down to zero before the government committed new funds for the banks. Today, we expect owners to intervene before that becomes necessary. Also, banks' capital positions have been greatly improved and our ability to spot a crisis has been enhanced through extensive surveillance work. Thus, the likelihood of a banking crisis is - hopefully - much less. But, should a new crisis emerge, we will expect the owners to act responsibly and stand by their bank with the full strength of their banking group - if required.
Role of the government
In a financial crisis, the government is responsible for the financial system as a whole. But, as we all know, how and when to call a potential crisis systemic is hard. Some will argue that it is difficult to define, but easy to spot once it is there. The Norwegian banking crisis became systemic when new loan loss estimates were about to wipe out the banks' capital base. When the crisis hit the three largest banks - at the same time - it was fairly obvious to us that the crisis had become systemic. At that point the Government just intervened to save the system. The need to save the system will be as strong today - if a new crisis should emerge. But it is not obvious that the same resolution methods will be applied
First of all, I think we should look harder for private solutions. This could also involve foreign investors. During the banking crisis, this option was closed due to existing banking regulations. Today, we have a more open attitude towards foreign ownership of banks. [In fact, almost 1/3 of our banking sector is foreign-owned.]
And secondly, we should not hesitate to close banks. During the banking crisis one small bank was closed and another small bank was wound down. This involved - as you may have experienced yourself - a lot of complicated legal issues. Today, we are better equipped to handle a bank closure. Payment system procedures are in place to handle banks with insufficient funds and we are continuously improving our systems to minimize the negative effects of a bank closure.
Some will argue that crisis resolution should continue to be wrapped in the misty veil of "constructive ambiguity". I do not support this view. On the contrary, transparency about our crisis resolution policies may reduce pressure for early government intervention. Ambiguity should, however, prevail about whom might receive support in a crisis. No institution should be given the impression that they are "too big to fail".
Role of the central bank
The lender of last resort policy has just been reviewed by our Board. They noted that deep and liquid markets now provide banks with a wide array of funding sources, and supplementary collateralized lending from Norges Bank has proven to be quite adequate so far. In fact, no emergency liquidity support has been required from Norges Bank since the last banking crisis. However, should a general liquidity shortage arise in the banking system as a whole, we will be prepared to supply the necessary liquidity to the system, if required against non-traditional forms of collateral.
In the event of liquidity problems in an individual institution, the Board confirmed that we will stick to the classic rules for lender of last resort. Support will be given to solvent, but illiquid banks against collateral at a penalty rate. The Board also stressed that the central bank's role as lender of last resort should be restricted to situations where financial stability may be threatened without such support. This is in line with what we have expressed in letters and speeches since the banking crisis.
The Norwegian banking crisis was a slowly moving solvency crisis. The crisis peaked several years after the first signs of banking problems. We had a fairly good picture of what was going on, but even then there were widely divergent views about what to do. If a banking crisis were to erupt again, I am afraid we may not get that much time. Liquidity problems could suddenly emerge in a large bank during mid-day settlement, which will only give us the evening to sort out the problems. Do we have a systemic situation? Is the bank solvent? These will indeed be hard questions to answer at short notice.
Should the central bank have its own view of the solvency of major banks, so that it can act quickly? Is such information at all necessary if the liquidity support is fully collateralized? Or will collateral typically not be available - at least in traditional form - in a crisis? And then, why bother with collateral if the liquidity support has to be "approved" by the Government? These are just some of the issues that we will address later today, and I look forward to this afternoon's session.
Summary and conclusions
Let me summarize.
How could this happen?
The Norwegian banking crisis was a boom-bust crisis with some special national features. With hindsight, the policy mix in the pre-crisis period was far from optimal. As a result, there were no effective barriers to the explosive growth in bank lending.
Could it happen again?
Not on the same scale, I would hope. Banks' capital cushions are much better today. We have better and more balanced macroeconomic policies. And, our flexible inflation targeting regime supports our twin goals of price and financial stability. Thus, the likelihood of a new banking crisis is low today.
How was the crisis resolved?
The Norwegian banking crisis was resolved quickly with little cost to the taxpayer. Private solutions were widely used in the first stage of the crisis. Banks' share capital was wiped out before any government funds were committed. The government did not provide a blanket guarantee, although general public statements were made to secure confidence in the Norwegian banking system. And, separate asset management companies were not established.
And, what can we learn from the crisis resolution methods adopted in Norway?
The resolution methods we adopted in Norway are now part of the standard toolbox for crisis resolution. We have learnt to:
* Explore private solutions first
* Expect more from shareholders
* Act swiftly to limit contagion
* Provide liquidity freely on good collateral.
Commit Government funds to save the financial system
Bearing in mind the gravity of the situation and the time available, we feel that the Norwegian banking crisis was handled quite well. Thus, I hope you will find that there is something to be learnt from the resolution methods we used.