Banking Soundness and Monetary Policy
Chapter

19 Loan Loss Recoveries and Debt Resolution Agencies: The Swedish Experience

Editor(s):
Charles Enoch, and J. Green
Published Date:
September 1997
Share
  • ShareShare
Show Summary Details
Author(s)
STEFAN INGVES and GÖRAN LIND 

There are many different causes behind bank failures, with losses in the credit portfolio being only one, albeit an important one. The theme for this paper is based on bank problems emanating from substantial loan losses of which a large share stems from the property sector.

In order to present a comprehensive picture of the process to recover loan losses in ailing banks, it is necessary to start at an early stage in the process—even before the bank is certain that a loss has occurred or has determined the size of it. First, one must assess the magnitude of both the individual potential loan losses to a bank and its aggregate potential loss. If several banks have simultaneous loan loss problems—the country is facing a systemic bank crisis—the total effect on the banking sector and the economy as a whole must be considered. The authorities must judge how best to solve the crisis. The closer the situation gets to a systemic crisis, the larger is the likelihood of state intervention, including state financial support. In the process, the state may assume ownership of banks or companies especially established to handle non- or low-performing loans and other assets. Figure 1 describes such a model.

Figure 1.Bank Restructuring: A Step-by-Step Process

Source: Arne Berggren, IMF Workshop on Bank Restructuring, 1996.

It is important to recognize that, irrespective of what support measures are applied, the loan losses have already occurred or are likely to occur in the immediate future. The first task therefore is to minimize the losses (and later to recoup as much as possible). This task is in itself very important and may save large sums. However, the recovery of losses should be seen in conjunction with the handling of the remaining “good parts” of the banks. Proper treatment of the problem assets will facilitate the future operations of the “good bank” and enable it to create profits again, which could compensate the state for earlier disbursements through future payments of dividends and taxes. Therefore, solutions to bank problems must encompass both the “good bank” and the “bad bank” (or, more accurately, the bad assets).

A proper valuation of the bad loans and other assets is important not only for the loan recovery process but also for general economic conditions. For instance, one often hears the argument that a bank (or the state, when assuming responsibility for the handling of a problem bank’s bad assets) should not liquidate a company since that would lead to large macroeconomic losses—such as increasing unemployment in an already underemployed geographical region. Thorough assessments of this company’s future development, including cash flow and earnings analyses, would present a fair picture of the financial cost of not liquidating. This could then be compared to the structural gains of keeping the company afloat. The state may still decide not to liquidate the company, for structural reasons, but this should not be considered as a measure within the bank support framework.

When the bad assets are ready for sale the state may face another macroeconomic problem: how will the sale affect the market in general? A typical example can be derived from the Swedish bank crisis. The state assumed ownership of large quantities of real estate holdings, which had formerly been in the banks’ hands as collateral to loans that had become nonperforming. The state aimed at selling these assets as soon as a reasonable price could be obtained. However, there was simultaneously a large overhang of unsold real estate held by the banks or other market participants. The market was very weak and was supposed to remain so for a number of years. The dilemma for the state was whether to maximize its own profits at the risk of ruining the real estate market for several years more.

Background to a Bank Crisis—The Swedish Case

This paper is built on the experience of the bank crisis in Sweden from 1991 to 1994. In my view, most of these experiences, including the methods and solutions adopted by the national authorities, are of a general nature and can be applied to other situations and other countries. Nevertheless, it should be noted that Sweden is a small country with a limited number of banks, which facilitated the handling of the crisis.

To better understand some of the choices that the authorities and the banks faced in the crisis a short background is helpful. This description has a sketchy character and covers only the most pertinent points.

During the 1980s, full deregulation of the Swedish financial markets took place. At the same time, there was a protracted economic boom that lead to high investment, in the real estate sector among others. Prevailing tax rules favored borrowing over saving. Further, because of earlier restrictions on borrowing, there was strong pent-up demand. Bank managers wanted to gain market shares and lowered credit standards, despite the fact they were lacking adequate knowledge and procedures to make proper credit assessments for new borrowers and new loan categories (such as lending to unfamiliar economic and geographical sectors). The volume of bank credits expanded rapidly from 1985 to 1990, and lending from banks and mortgage institutions increased from 85 to 135 percent of GDP (see Figure 2). A large part was lent to investors in housing or commercial real estate. Another part was devoted to other investments, most of it collateralized by real estate. Prices in the real estate market rose by 20 to 30 percent yearly during the latter half of the 1980s.

Figure 2.Total Lending

(In percentage of GDP)

Source: Bank of Sweden.

Around 1990, several factors caused the Swedish economy to deflate, hitting the real estate market particularly hard. Over an eighteen-month period, property prices plummeted by more than 50 percent in some locations. Since the cash flow was insufficient to pay interest on the loans, many property holders went bankrupt.

The situation worsened in 1992. A crisis for the Swedish currency, which was then linked to the European currency unit (ECU) basket of currencies, led to very high interest rates (peaking at 500 percent) and later to a substantial depreciation in the krona when it started to float. Both these events undermined the capability of many borrowers to service their debt. (It should be noted that many borrowers had taken loans in nondomestic currencies to benefit from their lower interest rates.)

The krona crisis also had more general repercussions for the Swedish banking sector and the Swedish economy as a whole. Foreign counterparties questioned the creditworthiness of Swedish banks and were reluctant to prolong existing credit lines necessary for short-term financing. Loans from the central bank could compensate for part of this loss but could not stem the substantial loss of confidence and the shrinkage of credit lines. Without swift and resolute action, there was a risk that the loss of confidence in the banking sector, not least by non-Swedish lenders, could spread to the government’s own substantial borrowing abroad. Thus, the government had to handle the double problem of a systemic crisis (five of the six major banks were affected) and a spreading loss of confidence. In response, the government issued a guarantee stating that no depositors, lenders, or other counterparties to Swedish banks—with the exception of equity holders—were to suffer any losses from their transactions with these banks. A few months later, this guarantee was confirmed by the parliament, which also formulated a general plan for the solution of the crisis. This plan provided, among other things, for unlimited use of public funds to finance support agreements between the government and the banks. The plan also included the establishment of a new agency under the Ministry of Finance, the Bank Support Authority (BSA), to handle all matters pertaining to support issues.

Against this background, these were the main considerations for the authorities:

  • the need to show resolute and quick action to restore confidence in Sweden and abroad;
  • the need to handle a crisis affecting almost all of the major banks; and
  • the need to handle substantial amounts of bad debts, a large part stemming from the collapse of the property market.

Bank Support Authority

Considering the severity of the problems and the need to restore confidence, it was a priority to start active support to the banking sector as soon as possible. Preliminary work could be done at the Ministry of Finance, but for such a wide-ranging crisis, its resources—such as staff numbers and skills—were not adequate. It was decided to set up an agency under, but separate from, the ministry. There was no time to build a large organization so the BSA started with a small number of employees and hired many outside consultants—mainly from abroad—with experience managing other crisis situations.

Banks requiring financial support from the BSA were asked to follow a certain procedure. After a preliminary application, the bank had to submit information for an assessment of its current situation (especially financial) and future prospects. Then the BSA would judge whether there was a need for support. If it was needed, more detailed information was required, including a comprehensive valuation of all bank assets—performing and nonperforming—together with a detailed assessment of future cash flows and profits from the bank’s ordinary operations. On this basis the BSA, after considering the views expressed in consultations with the central bank and the supervisory authority, took a preliminary decision, which was forwarded to the Ministry of Finance for final endorsement. The decision stipulated the form and conditions under which support would be provided.

The BSA operated on the basis of a number of general principles, particularly to minimize the use of state (taxpayers’) funds and to assure banks of equal treatment. However, since all banks were affected in different ways, some flexibility in the support measures was also necessary to achieve optimal results.

The Hammock Approach

A fundamental tenet guiding the support operations was the hammock approach (see Figure 3). All financial information obtained from a bank and from other sources, including macroeconomic data and forecasts, was fed into a forecasting model. This model could then produce a likely estimate of the bank’s financial development over the next three to five years.

Figure 3.Expected Capital Ratios: Different Scenarios

The result was used for different purposes. First, it was used to decide whether the bank would become profitable again within a reasonable timespan. If this was not the case—see Bank C in the graph—the bank would have to cease operations and be liquidated or merged. On the other hand, if the bank was considered likely to come close to the minimum legal capital ratio for a short period—as with Bank A—this would primarily be a matter for the present owners, maybe with some minor underpinning from the state, perhaps in the form of guarantees. Bank B illustrates the case where banks need substantial support to compensate for already incurred losses. However, to qualify as a B-bank the analysis must show that the bank was financially sound in its basic operations and would be profitable again in the medium term and thus had a viable future.

In addition to this analysis for the individual banks, the BSA also considered the structural implications of support for the banking sector as a whole. An explicit goal of the support operations was to forge a strong and competitive banking sector in Sweden (although fully open to access from abroad). This goal came into the forefront when one C-bank was to be merged with another bank. Since the state for financial reasons had been forced to assume ownership of the C-bank it was also responsible for the operation of selling the bank. When choosing the other party of the merger both financial and structural aspects were taken into consideration. Luckily, the solution finally found in this case was judged, by independent consultants, to be the best one on both scores.

As mentioned, the first step in the hammock approach is to decide whether a bank is viable for future profitable operations and thus for financial support. In the next stage the question is to decide how much financial support is needed. The balance here is between using a minimum of government funds while still providing the bank with adequate capital to enable fair competition with other banks. It should also be borne in mind that too little capital would risk short-term setbacks if the bank has to ask the BSA for more support. This would be politically embarrassing and could reduce confidence in the support operations of the BSA.

One way out of this dilemma is to provide part of the support in the form of guarantees. If the capital ratio falls below a certain threshold, the guarantee is converted into loans or equity capital. Another way of reducing the uncertainty concerning the bank’s future is to place the major part of its bad loans and assets with another company, not owned by the bank. That relieves the “good” bank of the volatility following from the value fluctuations of the bad assets and the costs of financing them. (This separation of assets is discussed later in this chapter.)

The hammock approach was used in a similar manner both when refinancing the good bank and when funding the company handling the bad assets. In both cases the principle of minimal but adequate financing applied. For an asset-managing company, the cash flow is highly negative in the first stages but turns gradually positive when asset sales start. Relatively large funding is necessary to bridge that gap.

The Hole

In order to refinance the problem bank to the level needed for future operations, the magnitude of losses—actual and probable—must be assessed. This quantity was termed “the hole” because the sum of the losses forms a hole in the balance sheet and capital of the bank. It is important to note that the hole is something that has already been incurred. The bank owners and the authorities cannot reduce it by neglecting it, only by acknowledging it and taking measures to contain it. Banks or countries that chose to hide the full extent of the hole have in reality suffered because uncertainty among investors and other counterparties leads to less favorable trading conditions, such as interest premiums on lending to the bank or country concerned.

Of course, the hole cannot be measured with a high degree of precision. It is an aggregate estimate based on a careful valuation of all the bank’s assets, performing and nonperforming. The valuation includes forecasts of the future behavior of borrowers, for instance the likelihood that they will not perform on their obligations to the bank and, if so, what the value of the collateral will be if collected. This leads to only a rough estimate of the hole and of the need for financial support. However, even a rough estimate is adequate in a crisis situation.

Measuring the Hole

In Sweden, a bank applying for support had to submit a very detailed valuation of all its assets. Special emphasis was placed on nonperforming and otherwise doubtful assets and their corresponding collateral. The aim was to assess the probable total loss amount, the hole. But the bank’s other assets were also valued in order to make an estimate of the bank’s future earning capacity.

Each loan was valued according to a seemingly simple and rather crude method: the risk of loss—within the next twelve months—was multiplied with the calculated value of the loss. For example, if the loss risk is 50 percent and the nominal amount of the loan is 100, the expected loss is 50 on the assumption that the whole loan will be lost. When there is collateral this was included in the calculations in the following way: the loss risk is still 50 percent, the loan amount 100, and the collateral is valued at 80. The calculation is now: “The risk amount” 20 (100 minus 80) multiplied by a risk factor of 50 percent, which results in an expected loss of 10. Since the analysis is built on probabilities, the outcome often will not be correct for each individual loan but should provide a reasonable estimate for the total portfolio.

In practice, however, the valuations were much more difficult and time consuming than indicated above. As a first step each bank had to compile a comprehensive list of all outstanding problem loans and assets. These loans and assets were then categorized and regrouped. (For example, different branches of the bank may have had claims on companies within the same company group. If this company group had financial problems, all claims should be valued together.) Since a major part of all the banks’ claims were related to property loans or loans collateralized by property, the BSA laid special emphasis on a fair and coherent valuation of property. A special property valuation board, composed of independent property experts, was established at the BSA to formulate valuation standards for the banks to apply. In order to ensure that the banks’ valuations conformed with market practices a large sample, about 25 percent, of the assets was also valued by independent market experts to get a “second opinion.” The valuation board at the BSA then further verified the accuracy of the banks’ valuations by duplicating the banks’ valuations on a small sample (approximately 5 percent). If the property valuation board found that a bank either over- or undervalued its assets, it had the right to inform the BSA which could then adjust the bank’s estimated amounts when assessing its financial situation and support needs.

The basic principle of the property valuation was to assess a market value. It should not reflect a “panic sales value,” which is far below a sales value under nonforced circumstances. Even so, it was often difficult to find a reasonable market value in a situation when the Swedish property market was more or less nonexistent. A supporting method of calculation was then added, based on the present and future cash flow generated by a specific piece of property. As a principle, the net yield (rent revenues less operating costs, including interest payments and taxes) should be in line with comparable properties in the market. The value of the estate was then assessed accordingly (after adding the yield premium an investor in properties will demand for not investing in risk-free assets).

Similar cash flow calculations were also used to value loans to industry and other businesses. Simply put, the valuation was based on calculating which amounts of write-offs and/or capital injections from the bank the borrower needed to return to a positive cash flow after servicing his or her financial obligations.

All the valuations were performed in accordance with a common—for all support-seeking banks—macroeconomic scenario supplied to them by the authorities. To achieve coherence, it was also important that all assessments referred to one specific day. After having “calibrated” the assessments in these ways, the BSA could then use different dates and economic scenarios to make sensitivity analyses of the data. It was also essential that all banks used the same underlying assumptions so support measures could be decided using comparable information.

Filling the Hole

The hole was filled in two ways: by recapitalizing the “good part” of the bank and by funding the separate entity to which the major part of the bad assets were transferred. In order to fulfill statutory capital requirements, a certain part of the state financial support had to be in the form of equity or equity-like instruments. The residual could be in the form of loans or guarantees, although guarantees do not improve the bank’s cash flow and are therefore less helpful. Loans, however, have the disadvantage of not providing a stake in the so-called upside. When the bank becomes profitable again the equity holders benefit from increasing equity values, but a lender will receive only the negotiated interest rate.

It is appropriate that the present equity holders lose their ownership of the bank in relation to the amount of the financial contributions from the state. There is no reason why owners whose bank has failed and who are not willing to make new investments in the bank should benefit from the expected increase in the value of the bank resulting from the state’s support measures. Such behavior would also be difficult to explain to the public-at-large.

The Asset Management Company

An asset management company (AMC) is the entity created to handle the problem loans and assets of a bank. AMCs existed in the United States long before the bank crisis in Sweden. Some specific features of the Swedish AMCs were the large share of property-related loans and assets in conjunction with a less liquid market. The character of the bank crisis as being systemic and having macroeconomic repercussions was also taken into consideration when forming the Swedish AMCs.

Definition of Terms

The term “debt resolution agency” is not always used consistently, and it may be helpful first to use two other terms:

  • Bank support authority: This could be operated by the state or by a private organization, such as a deposit guarantee fund. Usually, the task of the authority is to assess needs and to provide support to banks and to depositors. It may also be given the task to solve bank problems by facilitating structural changes such as mergers. But the authority should not, normally, be involved in sales of assets on a piecemeal basis because that might confuse its role and force it to build a large and heterogenous organization. The authority might, however, have the mandate to sell the whole asset portfolio of a bank to facilitate a quick reconstruction or merger.
  • Asset management company: This is a company to which the bad loans and assets are transferred from the problem bank. The AMC could be owned and funded by the state or by private owners. When restructuring severe bank problem cases the AMC must be totally separated from the parent bank, but in other cases the AMC might be a subsidiary or even a division within the bank. Irrespective of its organizational character, the AMC has a straightforward work plan. First to convert all loans into assets (real or financial). Then to restructure (“package”) these assets into easily sellable forms, and finally to sell them. The overriding goal of the AMC is to maximize the net sales value, after operational and financial costs, of its assets.

Arguments for Creating a Separate AMC

From a funding point of view, there is no restriction on keeping the bad loans and assets in the bank itself. Financial support from the state and other sources could fill the hole of earlier losses to enable the bank to continue its activities and to finance the holding of its bad assets. This is not a practical solution, however, if the amount of bad assets is large, and it may not provide the best economic result for several reasons.

First, the handling of bad loans and assets requires other skills than are normally available in a bank. Real estate specialists, liquidation experts, and people with insights into various industrial sectors may be needed.

Second, managing the bad assets would interfere with the daily running of the bank. One could easily envisage that important decisions regarding bad assets would receive lower priority than incoming new business to the bank. If not separated, the bank’s management and board of directors would have to use scarce time to discuss matters relating to bad assets instead of focusing on current business and strategy planning.

Third, both the good bank and the AMC could be given independent and transparent profit goals if separated. That would provide incentives for managers and staff.

Fourth, if the state owns the bank with the aim to sell it, a separation will facilitate a future sale of the bank. It will also probably fetch a higher sales price since the investors will have a clearer picture of the real value of the bank. In effect, the latter also applies when a separate AMC is funded by a bank that has not received state support.

Private or State-Owned AMC?

When a substantial amount of bad loans and assets has to be transferred to an AMC over a short period, it is normally impossible to find a private investor willing to own such an AMC without asking for far-reaching state guarantees covering the future value of the asset portfolio. The state is usually in a more favorable position owning the AMC itself rather than providing guarantees since it might then benefit from any future upward price movements (the so-called upside) of the AMC’s assets. In addition, it is difficult to formulate guarantees that would give a private owner strong incentives to sell the assets at the best prices. This could lead to further losses for the state.

Another option when a single bank has a medium to large amount of bad loans and assets is to remain within a subdivision or subsidiary of the same bank. In doing so, the bank would benefit from its knowledge and close contacts with borrowers. If the state has provided financial support to the bank, it may ask for a share in the “upside” of the AMC subsidiary.

Finally, a reasonable amount of small problem loans should remain in a bank’s ordinary organization, even if the bulk of bad assets are transferred to a separate AMC. Apart from the argument of “equal treatment,” there is a need to maintain the capability within a bank to handle workout procedures. In the Swedish case, all bad loans and assets with an assessed individual value under US$1 million remained in the problem banks and were not transferred to the AMC.

How Much Should Be Transferred to the AMC?

An important question is how large a portion of the bad assets to transfer from the good bank to the AMC. The reasoning here is structural rather than financial. If all bad assets were transferred the good bank would be in excellent condition and would surely have a strong future development. This is probably not an optimal solution, however, considering the competitive situation in the banking sector as a whole. Why should a bank be given this favorable position, especially since the same bank recently proved itself unsuccessful in its credit operations? Preferential treatment would also cause resentment in the other banks and among the public.

The Swedish authorities decided to leave the problem banks with a ratio of bad loans and assets similar to that of other banks in the market. As noted above, the bad assets remaining with the bank were mainly small ones, those that may best be handled in the bank itself since they do not require special expertise.

How Many AMCs?

We have no rigid views as to the number of AMCs. A very large AMC may obtain economies of scale but could also become unwieldy, which might hamper the ability to react swiftly, such as in sales transactions. Thus, while the number of bad loans to be handled by one AMC could be quite large, it should also be remembered that problem loans and assets require far more work than a similar number of ordinary loans.

Another factor to consider is the extent to which the impaired loans have been “processed.” If most of the loans in an AMC have reached a rather advanced stage, such as being close to the final sales, it may not be efficient to transfer to the same AMC a large number of fresh bad loans. Here, a solution might be to set up a parallel AMC and to merge the two AMCs at a later stage.

Different Stages of an AMC

AMCs will pass through the following six phases:

(1) Creation: “In the beginning there was chaos” says the Bible and this holds true in the case of Swedish bank restructuring. The AMC is normally created in a situation of crisis, general mistrust, and despondency. Rapid action is needed to avoid further financial losses, but the assets are not clearly identified and sorted properly, important documentation is lacking, and the organization is not in place.

(2) Start: At this stage, the authorities identify, document, and sort out the bad loans and assets, which form the basis for the AMC’s activities. They also establish an appropriate organizational structure, which will reflect and facilitate the work process.

(3) Construction: This phase is dominated by the handling of loans and other engagements. If this is found to be the most financially sound solution, negotiations with a borrower may result in a reconstruction. In other cases, loans will be transformed into assets (real estate, equity, or other), which are seized by the AMC. In some cases, bankruptcy is the only solution.

(4) Consolidation: Almost all holdings are now in the form of assets. Some are financial but most are real, such as industrial companies and real estate. The AMC is now restructuring and reorganizing its holdings to increase sales values.

(5) Dismantling: Most of the assets are now ready to be sold. The AMC is looking for buyers and is conducting sales negotiations.

(6) Repayment: The AMC’s outstanding loans and other obligations are honored, and any net worth is repaid to its owner.

These phases are discussed in greater depth below.

Creation Phase

Creation is a critical phase because important decisions must be taken under duress; politicians, the general public, and other market participants exert pressures on the process. The decision to establish the AMC, its general organization, activities, and goals must be clear, yet flexible. Sweden’s experience shows that during the life of the AMC shifting external circumstances may motivate changes.

The sole goal of the AMC should be to derive a maximum value out of the resources transferred to it. Such an unambiguous goal facilitates the work of the AMC and sends a clear message to politicians, the public, and market participants.

Starting the AMC

As soon as the AMC has started operations, the transfer of assets should begin. In Sweden, a basic principle was that the transfer prices should be set at fair values, reflecting market prices. For this purpose, the earlier careful valuation of a bank’s assets when applying for state support was useful, even if it had to be updated and supplemented. Using realistic transfer values fulfills several objectives:

  • It provides a better basis for the funding of the AMC. Normally, the funding of the AMC is such that when all assets are sold no money should be left. If the assets are transferred to the AMC at too high values, there will be a need for further write-downs and infusions of capital. This is usually politically inconvenient and may send negative signals to the market—perhaps that there are further losses “hidden” somewhere.
  • It provides a clear goal for the management and staff of the AMC. If they can return some of the original equity capital in the AMC, they have done a good job, assuming that market prices do not move in unexpected directions.
  • It provides clarity to the “good bank.” The good bank is financially compensated for the write-downs of the assets transferred (as well as for any remaining bad assets). This gives the good bank a fresh start.

It is important to emphasize that there are no financial gains to be derived from setting unrealistic (too high) values on the bad loans. The losses on these loans have already occurred—or are certain to occur—and they must be covered one way or another. By making the valuation fair, the authorities, the public, and other participants in the financial markets obtain a transparent picture of the situation. This helps restore confidence. The amount of support from the state is not dependent in the long run on the valuation of the assets transferred since it must still cover the real losses. Depending on the valuation, however, a larger or smaller portion will be provided to the good bank or to the AMC respectively.

The first handling of the bad loans and assets at the AMC consists of their thorough classification and documentation. Even though much work was done before the transfer, it is still necessary to sort the loans/assets into different categories to receive appropriate treatment. Usually, much of the necessary loan documentation is lacking and must be supplemented or reconstructed.

In this phase, the detailed organization of the AMC is built and staffed. It is often useful to shape an organization that reflects the structure of the various categories of holdings. In one Swedish AMC, there were subdivisions for industrial holdings and for real estate holdings. The subdivisions took the legal form of normal limited companies under a common parent company. The remaining loans and other liquid assets were held in another subsidiary to the parent, in this case a finance company. By dividing the AMC into independent companies, it was possible to increase efficiency by creating smaller and more homogenous units with distinct goals.

When organizing the AMC, one must pay special attention to incentives, because the activities of an AMC differ in many respects from those of other organizations. For instance, the ultimate goal of the AMC is to sell all its assets and to extinguish itself. Incentives, but ones not so generous as to upset the public and the political leaders and thus undermine the credibility of the AMC, must be set so management and staff strive for this goal instead of unnecessarily prolonging the life of the AMC (and their own jobs) by so-called empire building. Incentives, such as bonus payments, must also be set to encourage employees to sell or reorganize the assets so as to maximize net worth. A further category of incentive must be provided to induce employees, especially key staff, to stay in the organization as long as they are needed, instead of fleeing to safer jobs when the end of the AMC is approaching.

Construction Phase

The next step is to extract the actual value from the loans. This can be done in several ways. (In Sweden, this process was termed “to crunch” the loans.)

The most common method of transforming a loan is for the AMC to assume ownership of the collateral. (This method may be combined with the other two methods mentioned below.) Another option is liquidation or bankruptcy, which may be the only way to recover some money. Since these proceedings normally lead to a large reduction of values, they should be used as a last resort. However, the threat of liquidation or bankruptcy may be an effective ingredient in negotiations between the AMC and the borrower. It should also be noted that the procedure leading to liquidation or bankruptcy is costly and time consuming, especially in the midst of a banking crisis when administrators tend to be overburdened.

Finally, if the underlying, industrial or other, operations are found to be financially viable, a reconstruction may be negotiated. In theory, the terms of the reconstruction should be such that the borrower is able to generate a net future cash flow from the operation large enough to enable him or her to service the reconstructed and reduced debts. In practice, the negotiations may often result in more favorable terms for the borrower. For instance, when a borrower is leading his or her own company, the value of the company may be sharply reduced if he or she were to leave. To avoid this, the lender is willing “to pay a price.”

The negotiations between the AMC and the borrower is usually made much more complicated by the fact that several other counterparties to the borrower are affected simultaneously. Some are better positioned to recover their claims (such as those having more senior claims) than others who are less well positioned. These factors enter strongly into the negotiations and may threaten to destroy values by leading to less than optimal solutions. The AMC should establish a reputation of being fair toward other parties. Trust between negotiating parties will increase the chances of obtaining rational and financially better results.

Consolidation Phase

During consolidation, most loans are transformed into assets of different categories. The task is to make the assets financially viable and thus attractive for a buyer. A most common cause why the borrowers ran into financial problems was that the revenue-raising capacity of their assets was reduced for some reason (mismanagement, adverse market conditions, and so on). Thus the AMC has to perform a job similar to that of a “company doctor” who is consulted to identify the root of the problem and to propose—even implement—the necessary measures.

For an industrial company, the AMC may have to sell noncore activities and make other operations more efficient by reorganizing and reducing staff. For commercial property and residential homes, several measures may be adopted. At the outset, the AMC is usually holding geographically scattered pieces of property stemming from different claims. In some cases, the AMC is holding pieces of the same properties as other claimants. To facilitate the handling of a specific piece of property and to make maintenance work and services cheaper and more effective, the AMC should regroup its holdings to obtain synergies. This is achieved by swapping holdings with other parties and by purchasing complementary property. A simultaneous step is to renovate the properties in order to adapt them to current demand. Reducing the vacancy ratio is one of the most crucial factors in improving cash flow, and a better cash flow is pivotal for the market’s assessment of the value of the property.

In this phase the AMC needs to be staffed with specialists (or may hire outside experts) in various fields, such as particular fields of industry, the service sector, or real estate. Both economists and legal experts, as well as industry and property experts, will be engaged. An efficient method of work may be for each specific cluster of connecting assets to form a project group.

This group should be composed of the specialists needed in the specific case. The same experts could be involved in several different project groups simultaneously. There would also be an independent (nonexpert) chairman of each group to coordinate the work. Only in exceptional cases would disagreements within the groups be referred to higher management.

Dismantling Phase

In this phase, assets have reached a state that makes them suitable for sale. The major issue is when to sell in order to obtain the best possible price. The original goals for the Swedish AMCs were long term. The aim was to disinvest all assets within ten to fifteen years. The reasoning was the dismal state of the property market at the time of the establishment of the AMCs. A huge oversupply of property had developed and forecasts for future demand indicated that many years were necessary to restore balance. The AMC feared that selling property at a pace faster than “highly gradual” would result in another downward spiral of prices.

Two intertwined reasons have lead to a significant shortening of this long-term approach. First, the property market improved faster than expected, at least in some locations. Second, this made possible a change in the calculations consistent with the overriding goal of the AMC—to recover maximum values after including all costs, including financing costs.

The “new” calculations were founded on the following basis: If property prices is expected to climb gradually, what rate of yearly price increases are needed to compensate for the financial and other costs of holding onto the assets? This calculation must be performed on each individual asset, since they have different characteristics both on the price/revenue side and on the cost side. The computations often lead to clear-cut results—for instance, that the value would have to increase by 44 percent over the next five years to justify holding onto it. A sale at present prices would then be assessed in relation to the prospective price five years hence (gauged by the property market expert).

This highly simple and pedagogical method facilitated taking and explaining many decisions leading to more rapid sales. In practice, it led to a radically altered time frame for the dismantling of the AMC. All assets will now be sold within five years from the start of the AMC. Admittedly, at the bottom of the AMC’s “cesspool,” there may be some assets having very low probabilities of recovering any significant amounts. Such assets could be bundled together and sold to an interested party, such as a debt-collecting agency, at a low price and without any guarantees from the AMC.

Apart from the economic rationale of shortening the lifespan of the AMC and the state’s asset holdings, a faster sales pace also has favorable structural effects by reintroducing the assets into the market. It is also a relief politically for the government not to be further burdened with financial responsibility for the AMC.

In the period of selling its assets, the AMC must endeavor to create trust with other parties in the property market. An AMC financially supported by the government may easily, even if undeservedly, attract a reputation for underselling—at too low prices—since losses will be compensated for. Such a reputation will cause ill will and may destabilize the market by increasing uncertainty about the AMC’s goals. The way to avoid this is to formulate a distinct goal of profit maximization and to give the management and the staff of the AMC incentives to adhere to that goal.

Repayments Phase

During the earlier phase of asset sales, the need for external financing of the AMC was gradually reduced and some loans and other subordinated debt were repaid. The final liquidation of the AMC implies transforming all assets into cash and terminating the organization and all its obligations. It is crucial that all steps be well documented to avoid future lawsuits and claims on the owners of the AMC.

The net worth of the AMC would be returned to its owners. Financially, it should be treated as a balancing item in relation to the original outlay. This means, for instance, that if government support to the banks has been financed out of a separate fund, such as a deposit guarantee fund, an equivalent (maybe with interest added) sum should be repaid to that fund.

Selling the Good Bank

As mentioned earlier, the process of recovering financial support to banks is not limited to recouping loan losses. An integral part of the process is the development of the “good bank.” If the support measures have been successful, the restructured and trimmed bank should, after some time, be ready once again to generate profits, taxes, and dividends.

In the case where the state has acquired a share of the bank, corresponding to the extent of its financial contributions, it will benefit. When the equity of the bank rises in value due to its increased capital and improved prospects, the state can sell all or part of its holdings. Although the state may face a political dilemma when deciding whether to receive dividends or to sell its shares, the benefits are equal in economic terms: a rational share price would be based on the discounted future income stream from the bank to the owner. Therefore, the share price and the future stream of dividends would represent similar values.

In principle, the aim of the sale should be only to obtain the best price possible (given that the new owners fulfill the “fit and proper tests” and other prudential criteria). The state may have additional goals, such as promoting structural changes in the banking sector by merging banks that complement each other. If so, the “cost” (the loss of revenue by not choosing the highest sales price) should be presented in a transparent manner to facilitate a comparison to the alleged structural gains.

The state may choose to disinvest its share holdings in gradual steps or to sell the whole bank at one occasion. If the goal is profit maximization, one owner may be willing to pay a premium to gain control over the bank. On the other hand, if a single owner cannot be found, trying to sell a large bank all at once could depress share prices. If the decision has structural implications, the government might sell the whole bank to another financial institution to achieve synergies; note, however, the argument about making the “cost” clear for all to see. If political considerations are important, it may be necessary to obtain a political majority for reducing the state’s holdings at faster than a gradual pace.

Irrespective of the arguments and alternatives chosen the sales process should fully conform to market principles, especially openness and access to information for all interested parties. If the state remains a large shareholder, it should not ask any favors not provided to other shareholders.

Actual Loan Loss Recoveries in the Swedish Case

The sum of all forms of state support to the banking sector, excluding guarantees that were not utilized, amounted to SKr 65 billion, or 4 percent of Swedish annual GDP. The latest forecast is that the government will recover at least SKr 48 billion of this amount. The recoveries will take the following forms:

  • SKr 6 billion from the sale of 34.5 percent of the equity capital of Nordbanken. The state became the sole owner of Nordbanken in the crisis;
  • SKr 5 billion from the redemption of some 5 percent of the equity capital of Nordbanken. The bank used excess funds to buy back a part of the government’s shares in order to cancel them;
  • SKr 17 billion from the sale of the remaining shares of Nordbanken (based on present share values quoted on the Stockholm stock exchange);
  • SKr 5 billion from dividends on the government’s shares in Nordbanken;
  • SKr 11 billion from the liquidation of the state-owned AMC, when all assets are sold; and
  • SKr 4 billion already recovered from the merger of the two original AMCs. At the merger, the first AMC paid SKr 4 billion to the state to acquire the assets of the second AMC.

It should be noted that in an economic sense the above calculations paint too rosy a picture since they do not account for the time value of money, as measured by the financing cost of the state support. Partly to compensate for this factor, it would not be appropriate to count tax revenues to the state from the reborn banks as “funds recovered.”

The recovery rate for Sweden is only illustrative and not an indication of a normal recovery ratio. Several developments contributed to make the ratio better than estimated at the peak of the crisis. First, the quick upturn of the Swedish economy following the depreciation of the krona and the fall of interest rates; this led to an improved situation for most borrowers, thus reducing the amount of nonperforming loans. The AMC could sell many of its industrial holdings more quickly than expected and receive better-than-expected prices. Second, the real estate market stabilized and turned upward, especially in attractive locations. Most of the real estate holdings of the AMC could be sold at or above their previously assessed values. Third, confidence in the banking system was rapidly restored, and banks could resume their ordinary business, which even during the crisis remained efficient and profitable. The banks also rationalized their operations and reduced their staff, obtaining significant cost reductions.

Important Lessons

What are the most important conclusions to be drawn from the Swedish experience with regard to minimizing the cost to the banking system and to the state budget?

  • Confidence in the system must be restored rapidly. Distinct decisions must be taken early and quickly transmitted to the market, as well as abroad. Any remaining uncertainty will affect the banks in various ways, leading to increased losses in their operations.
  • The solutions should not focus solely on the loan losses. In addition to recovering loan losses, large amounts may be gained or lost by restructuring the “good part” of the bank. A rapid and thorough investigation of the total situation of the bank is necessary. After that, it must be decided whether the bank has a viable future and what support is needed. The bad loans and assets must be taken care of anyway, and a higher price for large quantities of assets will typically be obtained if they are first “processed” and later sold through an AMC. Smaller quantities of assets may remain in the bank or be sold directly to another party.
  • The hammock approach was found valuable for deciding on the amount of funding to the good bank and the AMC. State-supported funding should be adequate but not excessive. When choosing between different forms of state financial support, forms that enable the state to benefit from expected future increases in the value of the bank should be given prominence.
  • Valuation of loans and other assets should be careful and follow strict and equal rules for all banks. Transferral of assets to an AMC should be at market prices.
  • The transfer of assets to separately established AMCs proved appropriate both from an organizational and from a financial point of view.
  • The AMC should have a strict profit-maximizing goal. In all its operations, it should act as a normal market participant.

This paper has focused on issues at the “micro level” the macro aspects have not, purposely, received due attention. It should, however, be made clear that there is a close mutual interdependence between the restoration of stability in the financial sector and a sound macro-economic development. Costly state intervention to save banks and to stabilize asset values may well be wasted if the banks are not supported by appropriate macroeconomic policy measures. The reverse is also true: problems in the financial sector will undermine overall economic stability and growth.

The authors wish to express their gratitude to those persons who have contributed their valuable insights to this paper, especially Hans Jacobsson, who succeeded Stefan Ingves as director general at the Bank Support Authority, and Jan Engstron at the workout company Securum. Samuel Coleman, managing director of Salomon Brothers, Inc., has shared with us his unique knowledge of Swedish banking conditions combined with experiences from other countries. This being said, the authors assume full responsibility for the text presented.

    Other Resources Citing This Publication