Risk Management, Expectations and Global Finance

1 Introduction

The banking sector underwent fundamental transformation, beginning in the early 1970s. Over the next decades, most countries deregulated their financial markets and lowered entry barriers for foreign investors. Cross-border flows increased rapidly, while new securitized assets and derivatives (mortgage-backed securities, exchange swaps, zero bonds) emerged.^[1] Digital trading made financial transactions not only faster, but also less expensive. To a certain extent, this allowed households to participate directly in capital markets. At the same time, the financial industry became more heterogenous. Insurance companies, mutual funds, and postal and saving banks started offering their customers investment services, challenging the position of established commercial banks. Even though the share of finance relative to GDP has grown over-proportionately since the 1970s, many credit banks faced falling profit margins due to stagnating or even declining interest rate income. As a result, most institutions expanded their investment activities, while traditional lending became less important. German universal banks such as Dresdner, Deutsche and Commerzbank abandoned their mixed business approach with the aim of becoming global investment houses.^[2]

This transformation of the business model has been identified as one of the main reasons for excessive risk-taking by banks, leading to more bank failures and systemic financial crises over the past decades. According to this view, German banks – once the symbol of financial solidity – adopted the aggressive style of Anglo-American finance without preparing themselves for the specific risks and imponderabilities of these practices. German bankers – raised in the well-protected climate of Rhenish capitalism – were not familiar with the dynamics of global capital markets and the unwritten rules that characterized the financial centres of London, New York, and Singapore.^[3]

However, beyond anecdotal evidence, there is little empirical research on the risk-taking of financial institutions. Most studies in this field have focused on individual investors, while banks and other intermediaries have received scant attention. This is partly due to the behavioural turn in finance with its emphasis on cognitive biases and information errors at the micro level. The changing institutional environment and the specific forms of corporate risk management have been widely ignored, even though institutional investors have increased their share of financial transactions.

This article takes a more structural look at the transformation of risk-taking in the banking sector. Its theoretical background is the concept of financial intermediation. According to older versions of this theory, the main function of banks consists in channelling funds from household deposits to companies and investors who have different maturity and risk preferences. Banks play an important role in reducing transaction costs and information asymmetries.^[4] However, as Allen and Santomero argued, this traditional function of banks has become less important since the 1970s.^[5] Empirical evidence proves that financial intermediation has increased, even though transaction costs and asymmetric information have declined.^[6] The authors explain this development in terms of the growing costs of risk management. Especially in the first years of globalization, new financial products involved high risks, while the overall monitoring of markets by public media, financial analysts and rating agencies was not fully functional. At the same time, the supervision of banks by government departments, central banks and regulatory agencies became less strict. Deregulation lowered transaction costs on financial markets but increased the cost of risk prevention. Private households and small investors often lacked information about security markets, and they were unable to operate comprehensive risk strategies such as portfolio diversification or hedging.^[7] As a result, they frequently turned to bank services for their financial investments. Indeed, the share of equity owned directly by private households fell in most advanced economies during the period under consideration. While in 1970, around 80 per cent of all US corporate equity was held by individuals, this was lower than 50 per cent in 1985.^[8] This paper argues that commercial banks regarded the trading and management of financial risk as the core of a new business strategy. They heavily invested in their information resources and diversified risks both in geographical and temporal terms.

But how efficient was the risk management of banks in practice? Were financial intermediaries able to compensate for market frictions and the information biases of individual investors? Did they possess better information about complex financial innovations and international markets? How did banks assess long-term risks, and how much did they learn from experience? All these questions point to the larger problem of expectation formation, a problem which has been intensely discussed as a key variable in understanding financial markets.^[9] Especially behavioural economists have contributed to this debate, analysing the impact of past experiences on investment decisions. Drawing on the older theory of Kanneman and Tversky,^[10] Gennaiolo, Shleifer and Vishny introduced the model of diagnostic expectations. According to this model, investors rely on incoming information in making decisions, but overrate their direct experience, which they often consider to be representative.^[11] Other studies have shown that financial actors frequently operate with an extrapolation bias, thus overemphasizing recent experiences.^[12] Based on historical household surveys, Malmendier and Nagel have instead argued that long-lasting experiences – either biographical or as the result of collective memory – are more detrimental for investor choices than current (or very recent) information.^[13] The experience effect of financial shocks is particularly strong and explains generational variations in beliefs. Like most behavioural economists, Malmendier and Nagel question Bayesian learning and other rational strategies of agents to collect and assess information from other sources.

However, in all these studies it remains unclear what distinguishes the expectation formation of financial institutions from that of individual investors. More recently, Malmendier and Wachter have claimed that experience effects are not limited to untrained individual investors but similarly influence financial experts.^[14] In another study, Bouwman and Malmendier found evidence that banks who went through difficult times tend to build up higher capital buffers and apply more careful lending practices.^[15] There are reasons to assume that experiences affect the decision-making of financial institutions and account for macro-level phenomena such as asset pricing or corporate risk strategies.^[16] However, we still do not fully understand how banks use their knowledge from the past, for example whether this is based on institutional learning or on the personal experiences of bank managers. Moreover, it remains unclear whether historical learning is a linear process, given that lending behaviour of banks is highly procyclical due to biased memory effects.^[17] Finally, there is no overall evidence whether experiences from past crises are positively correlated with more financial stability.

The following article analyses the implementation of risk management by Germany’s largest financial institution, Deutsche Bank.^[18] Founded in 1871 to fund German industry’s foreign trade, Deutsche Bank soon transformed itself into a universal bank and offered a wide range of credit and investment services. Following World War I, the bank expanded into the home market and built up a network of local branches. Its international business continued to be substantial, but the domestic market gained in importance. Since the 1970s, however, the model of Rhenish capitalism with its close co-operation between banks, state and industry has fallen into crisis. Confronted with falling market shares and profit margins, Deutsche Bank thus decided to strengthen its foreign capital investments. It opened new branches abroad, expanded its global network of correspondent banks and acquired subsidiaries in financial centres such as London, New York, and Hong Kong. However, its transformation into a global investment bank, initiated in the early 1980s under the guidance of Alfred Herrhausen, took longer than expected. Internal struggles within management overlapped with structural limitations such as the lack of skilled investment bankers and the complex organization of the bank. In the first period of financial globalization, therefore, Deutsche Bank focused on segments where it already had long-standing expertise: the issuing of international bonds on the Euromarkets and the financing of sovereign debt in developing countries and the Soviet empire. This paper argues that these activities were detrimental to Deutsche Bank’s refined risk management system as it came to emerge in the 1970s.

2 Hazard and Crisis: From Gentleman Banking to Regulation

Like most banks, Deutsche Bank had not operated a centralized system of risk monitoring for a long time. Until the 1930s, the bank based its risk assessments on „bedrock principles“, such as credit limits and the requirements of liabilities for lending.^[19] Moreover, it kept a certain amount in reserves or liquid assets, and diversified credit risks across different branches and maturities. To finance large projects, the bank often formed syndicates with other houses to share the risks. In the case of very large loans, a more detailed assessment of balance sheets, business reports and the financial solidity of the borrower was required, even though such information was often less decisive than the reputation of the client.

While this system of informal rules was sufficient in times of stability, the situation changed after the First World War. Hyperinflation, financial volatility, and political crises led to a series of bankruptcies, while financial fraud and mismanagement increased.^[20] Just as many corporate clients became insolvent during the Great Depression, banks were likewise exposed to systemic risks. This was apparent after the banking crisis of 1931, when the bankruptcy of the Darmstädter Nationalbank triggered a bank run and forced the Reichsbank to bail out the major German credit houses.^[21] A series of legal reforms established a comprehensive framework of supervision and forced banks to introduce more formalized internal rules. The Banking Act of 1934 (Kreditwesengesetz) created a Banking Supervisory Office at the Reichsbank. The law introduced a formal system of licensing, rules on reporting, as well as reserve and liquidity requirements for commercial banks. Deutsche Bank responded by significantly expanding its internal accounting system. In doing so, it was complying with the new regulations, but it was also aiming to strengthen its own risk prevention. Although Deutsche Bank – which had merged with Deutsche Disconto Gesellschaft in 1929 – came through the crisis better than other major banks, it suffered heavy losses, and therefore learned lessons from the biggest financial crisis in German history.

The Banking Act of 1934, which provides the legal framework for banking supervision in Germany to the present day, not only passively exposed banks to state regulation, but made them regulatory actors themselves.^[22] The German tradition of corporate self-regulation continued to play an important part in the financial industry during the Nazi period, as well as in the postwar decades. Deutsche Bank thus expanded its central credit department and tasked the Supervisory Board with independently reviewing all large loans of RM 1 million or more.^[23] The bank also strengthened its economic expertise in anticipation of fluctuations and shocks. In 1931, an Academic Advisory Board was set up under the direction of the monetary expert Melchior Palyi, a professor of economics and director of the Berlin Institute for Monetary Research. In addition, the archive of the bank, which until then had mainly collected the annual reports of other banks and companies, along with evaluations in the economic press, was transformed into the Economics Department. A group of economists, statisticians and accountants now regularly prepared comprehensive reports on the economic situation, based on their own research as well as forecasts from other German research institutes.^[24]

Over the decades that followed, however, the Economics Department remained a relatively small unit within the bank and had a limited impact on its operational risk management. Under the Nazi regime, war funding became the major task of banks, while financial markets were repressed. As all economic and financial resources had to be mobilized for rearmament, the Nazis suspended capital and liquidity requirements for banks.^[25] After 1945, the bank was decentralized, and risks were managed mainly at the level of branches and the ten head offices in Düsseldorf, Frankfurt, Munich etc. When the bank was reunified in 1957, the decentralized structure partly survived, which made it difficult to implement a coherent and overarching risk system. Formalized risk control was, however, not a priority, given that the house pursued a rather conservative style of finance under the guidance of Herrmann J. Abs.^[26] Even after the return to free currency exchange in 1958, Deutsche Bank did not lend to foreign non-bank customers or make large investments in security markets.^[27] Nor did the house actively expand through foreign branches. It conducted international activities in a restrained fashion, based on a correspondent network, banking consortia and minor subsidiaries.

3 New Risks, New Rules?

This situation changed profoundly throughout the following years, as Deutsche Bank slowly re-entered international financial markets. It began in 1965 to grant direct loans to foreign non-bank clients and in 1968 to forfeit exports, which meant that the bank assumed the default risks of trade bills.^[28] In 1970, Deutsche Bank opened a subsidiary in Luxembourg that processed most of the loans to foreign non-bank customers.^[29] Over the next few years, further branches and subsidiaries were opened in London, Paris, New York, Singapore and Hong Kong as well as in emerging offshore centres such as Bahrain and the Cayman Islands. With the end of the Bretton Woods system of fixed exchange rates, currency risks increased considerably, boosting the demand for new hedging instruments such as foreign currency options, spots, or forward contracts. Proprietary trading on security and exchange markets became a highly profitable source of income. Moreover, Deutsche Bank established itself as the leading international issuing bank, occupying a strong position especially in the Eurodollar market. This market had emerged in the 1950s when American banks opened dollar-denominated accounts at British – and later at other European – Banks to circumvent the supervision of the US Federal Reserve Board.^[30] The Eurodollar accounts fuelled the emergence of a rapidly growing capital market. Deutsche Bank was particularly engaged in the issuance of large international bonds, often as a leader of bank consortia. In 1975, Deutsche Bank’s share of Eurobonds traded worldwide accounted for more than 12 per cent.^[31] The bank also became heavily involved in redirecting petrodollars from the oil-exporting countries into global financial circuits. By the end of the 1970s, international business represented more than 40 per cent of Deutsche Bank’s activities.^[32]

Within the banks’ management, awareness of rising international exposure grew. The almost entirely unregulated Eurodollar market with its overheated liquidity was seen as a particular danger, especially as this market was largely based on short-term interbank transactions.^[33] Bank-to-bank payments under volatile exchange rates engendered serious risks, as the failure of Herstatt Bank in June 1974 made clear when thousands of depositors lost their savings. Herstatt, a private investment house based in Cologne, had invested high sums in foreign exchange markets. In summer 1974 it had accumulated losses of 64 million DM, which consumed 89 per cent of the bank’s equity. The Herstatt crisis set off alarm bells on the management floors of Deutsche Bank, given that other German institutions such as the Hessische Landesbank and the Westdeutsche Landesbank had also suffered large losses. In June 1974, the major German credit institutes rejected the plan of a rescue of Herstatt, as proposed by the then president of the Bundesbank Karl Klasing. However, it was evident that the problem of increasing exchange rate risk affected the entire financial industry. Internal reports of Deutsche Bank referred to a „decline in morals and customs on foreign exchange markets“, something which was considered „incompatible with the conservative values“ of Deutsche Bank. More than that, it was feared that „speculative activity on this scale may sooner or later give rise to administrative countermeasures and must damage the reputation of the banks as a whole“.^[34] For this reason, the bank – at least for some years – abstained from foreign exchange speculations by external brokers and did not participate in floating-rate bonds and roll-over credits, which became popular in the 1970s due to high inflation.^[35]

As expected, the bank crash of 1974 led to increased pressure for stricter regulation.^[36] Already in September 1974, the Banking Supervisory Board introduced new rules for foreign exchange trading.^[37] In order to prevent further legal interference, German commercial banks founded a special institute, the Liquiditäts-Konsortialbank GmbH in Frankfurt. It had the task of providing commercial banks with liquidity in the case of payment defaults. Private bailouts had been unusual in the past, but they were now regarded as the means of choice to prevent a bank collapse. In line with this, the German Banking Association introduced a Deposit Protection Fund in November 1975. Almost all private banks contributed a fixed sum (one-third of the liable equity) to the fund in order to compensate depositors in case of a bank failure. Germany was thus a good example of a country where formal supervision by the state was partly supplanted by more regulation by banks themselves.^[38]

At the same time, the demand for international co-ordination and harmonization of supervision grew. While national legal standards still restricted transactions across borders, offshore finance and non-regulated interbank flows had amplified, with unpredictable effects on stability and risks.^[39] In 1975, the G-10 member states founded the Basel Committee on Banking Supervision under the auspices of the Bank for International Settlements.^[40] It enforced the monitoring of international financial markets and formulated some basic rules for the harmonization of supervision. These were just recommendations which were not binding on the member states, but they enhanced the awareness of new financial risks, for example, regarding the rapidly expanding Eurodollar market which was mainly based on wholesale banking transactions. The management of Deutsche Bank took the dangers of financial globalization seriously, as several internal reports on the Eurodollar market document.^[41] However, demands for greater government regulation were rejected.^[42] The bank’s line was that the financial institutes themselves would have to take sufficient precautions to contain new risks effectively.

Indeed, Deutsche Bank began at this point to reorganize its internal accounting and risk management. Internal credit monitoring became not only stricter, but also more centralized. Until then, corporate loans had usually been granted by branches, but central offices and the Management Board (including the credit committee on the Supervisory Board) now became more directly involved in controlling large loans. The bank also started to monitor the debt condition of large corporate clients independently of specific loan requests,^[43] and the bank’s foreign risk exposure was carefully monitored.^[44] In the following years, it introduced a new system of accounting and information flows, partly based on computerized data analysis. The Economics Department received new analytical resources and became more directly involved in operative and strategic decisions. In 1971, Franz-Josef Trouvain, a trained economist with a Ph.D. from the University of Marburg, was appointed director of the department. Trouvain had joined the bank in 1953 as personal assistant to the Board’s spokesman Hermann-Joseph Abs and worked in the Economics Department for many years. He now received the title of Chief Economist and cooperated closely with Alfred Herrhausen, who had joined the Management Board in 1971. Herrhausen, an economist himself, regularly used reports and statistics from the Economics Department for both public speeches and internal statements.^[45]

Herrhausen and Trouvain agreed that a more professional economic research unit was indispensable for a leading financial institution such as Deutsche Bank. They increased staffing resources and reorganized the department, which was subdivided into six units with research competences on the general economic situation, foreign markets, different branches, etc. The department was also responsible for the bank’s other information resources, including its library and archive.^[46] In 1977, the bank centralized its archive which until then had been scattered in different places. The archive’s main function was not to preserve historical documents – a separate historical archive had been founded for this purpose in 1961 – but to provide an up-to-date information system on business partners, foreign markets, and statistical data.^[47] A major task was to use evidence from the past for the new system of risk assessment. As we will see, this historical data would also be made use of in developing an internal country rating over the next years, in order to help evaluate the creditworthiness of sovereign debtors and make reliable predictions on inherent default risks. The reference to historical data reflected a general trend in economic forecasting where time series were used to calculate model parameters. While this approach had already been established in professional economic forecasting models beginning in the 1950s, banks only now began to analyse historical data thoroughly. While in Deutsche Bank, information from the archive had been used in the past for specific tasks, for example to provide information on the creditworthiness of business partners, it was now systematically integrated into the banks’ risk assessments. The Economics Department assembled data from a variety of sources – both external institutions such as the Deutsche Bundesbank or the OECD, and internal accounts – and was more proactive in its research. While previously, it had drafted statements mainly in response to demands from Board members, it now regularly reported on international banking, financial markets, or the macroeconomic situation in different world regions. Trouvain himself also played a major role in the professionalization of corporate scientific counselling and banking research in Germany, for example as the chairman of the German Association of Economic Councils within the Confederation of Employers’ Associations (BDA).

4 From Poland to Mexico: Banks and Sovereign Debt

Starting in the 1970s, financing sovereign debt became highly profitable for commercial banks.^[48] While public debt had declined (or at least remained stable) in most countries during the post-war era, many governments in Europe and North America – but especially in developing countries – now began to face increasing budget and balance of payments deficits.^[49] Structural changes such as the expansion of welfare expenditure went hand in hand with the increasing cost of oil, economic stagnation, and unemployment. Financing this debt via the capital market became more difficult, as soaring inflation made government bonds with long maturities unattractive. Many countries therefore introduced bonds with shorter maturities or inflation-indexed bonds.^[50] Such strategies, however, required a developed domestic capital market and a well-established banking system. Poor countries in the Global South (especially those without oil reserves) and many socialist states became increasingly dependent on foreign loans which were mostly provided by large commercial banks from Western Europe, Japan, or the United States.

Since the late 1950s, Deutsche Bank had regularly placed bonds of foreign municipalities, governments, and public corporations onto the German capital market. The bank also took a leading role in issuing large bonds for international development agencies such as the World Bank or the Inter-American Development Bank.^[51] In 1968, sovereign bonds accounted for 75 per cent of the bank’s issuance volume, while corporate bonds represented only 25 per cent.^[52] The bank considered the risk of such bonds manageable, even though in some cases the securities could not be entirely placed on the capital market. Deutsche Bank then had to temporarily put them onto its own books, often below par.^[53]

Until the late 1960s, banks had mainly funded the sovereign debt of OECD countries. This now changed profoundly. Deutsche Bank became involved in large credit operations in the Soviet bloc. In 1970, it was the lead bank in a German consortium that funded a 1.2 DM billion gas pipeline project in the Soviet Union. Even though the conditions of this credit were not very attractive, given a maturity of 12 years with an interest rate of only 6.25 per cent, Deutsche Bank was eager to lead the consortium for strategic reasons, as this credit looked like path to further credits.^[54] In fact, by 1978, three more large loans with similar terms had been granted to the Soviet Union under the direction of Deutsche Bank. In 1973, the bank opened an office in Moscow to enhance its position as a major financial partner of the USSR and other COMECON states.^[55]

While the Soviet Union had been able to improve its international payment position in the late 1970s due to its high oil revenues, the other COMECON states had come under increasing financial trouble. Since the early 1970s, Deutsche Bank had constantly expanded its credit lines, especially to Poland and the GDR, and offered these countries various facilities such as Eurocredits, medium-term investment loans, trade financing through forfaiting, or even the issuance of long-term bonds. Poland had already drawn on Eurocredits of around $500 million in 1975 and was thus only just behind the Soviet Union with a volume of $750 million. Deutsche Bank’s lending policy was based not least on political considerations. Since the early 1970s, economic relations with Poland and the GDR had been an important element of Bonn’s policy toward the East and were promoted by government measures. In 1973, for example, the federal government promoted loans to Poland with an interest subsidy, while the GDR was granted an interest-free overdraft (Swing) of DM 600 million to fund trade between the two German states.^[56]

However, against the background of a global economic recession, the foreign debts of the COMECON states reached dramatic levels in the 1970s. Between 1971 and 1981, they increased from $8 billion to $95 billion, while annual debt service grew from $1.6 billion to $24 billion. These were alarming figures in view of the high foreign trade deficits, which amounted to $4 billion for the entire COMECON area in 1981.^[57] High interest rates aggravated the problems of the debtor states. In Poland, the economic situation deteriorated during the political crisis in 1981 when GDP dropped by 13 per cent, while the country accounted for 30 per cent of all foreign debt in the COMECON states.^[58] Deutsche Bank’s economic research department had already warned about Poland’s impending insolvency in 1980 and placed the country in the worst group in its internal risk assessment. At the end of 1981, Poland’s foreign debt of $23 billion put it in third place among the group of less developed economies, after Brazil and Mexico.^[59] The bank also had substantial exposures to the other countries of Eastern Europe, especially Yugoslavia and Romania. A further reason for concern lay in the fact that COMECON no longer formed a unit of borrowers. Each country now had to cover its own liabilities. The umbrella theory of the 1970s had led many Western banks and governments not to examine the creditworthiness of individual countries too rigorously, expecting that the USSR would step in as lender of last resort. Against the background of the growing disintegration of the Eastern bloc, however, this assumption had become questionable.^[60] The debt crisis in Eastern Europe was also critical for Deutsche Bank because a large part of its loans to the COMECON countries had been directed to the state banks of these countries. When Poland suspended payments to foreign creditors in 1981, Deutsche Bank had an exposure of around half a billion DM in this country, for which adjustments were necessary.^[61]

Poland’s liquidity problems marked the beginning of a global debt crisis that quickly affected the developing countries of the Global South.^[62] The reasons were similar almost everywhere. Many poor countries had become heavily indebted in the early 1970s, when surplus liquidity from oil-producing countries entered the international financial system and capital became available on favourable terms.^[63] Low interest rates and lax lending by European and American banks played an important role in this. Banks granted large loans to developing countries in Asia, Africa, and Latin America, which had high growth rates and were on the threshold of industrialization.^[64] However, many of these countries ran into payment difficulties in the early 1980s when interest rates rose, and energy prices slowed economic growth after the second oil price shock in 1979. Especially countries without oil reserves now faced insoluble difficulties. Between 1971 and 1981, the gross debt of developing countries had risen from $87 billion to $524 billion, and debt service in 1981 amounted to $112 billion.^[65] Among the defaulting countries figured African states such as Tanzania, Ghana, Libya, and Zaire who urgently required debt rescheduling. Turkey and many Eastern European countries were also at risk. Even more serious, however, was the situation in Latin America, namely in Mexico, Brazil, and Argentina. Given their size and international exposure, a sovereign debt default would represent a systemic risk for global finance. Like many other European and North American institutes, Deutsche Bank had expanded its credit lines in Latin America since the early 1970s, although with lower rates than other international banks. In November 1977, for example, the bank participated in an international consortium with 113 other financial institutions that arranged a loan of $1.2 billion for Mexico.^[66]

Already in the wake of the Polish crisis, Deutsche Bank had started to build reserves. In retrospect, the high write-downs in Eastern Europe proved to be an advantage for the bank, as it began earlier than other institutes to reduce its international exposure. At the end of 1980, the Board of Management decided to transfer a "substantial part of the operating profit to risk provisioning."^[67] However, it soon became apparent that the crisis in Latin America would demand value adjustments in different dimensions. In October 1983, the bank had an exposure of DM 7.6 billion in 27 highly indebted developing countries. Brazil (DM 1.7 billion), Mexico (DM 1.1 billion), and Argentina (DM 0.8 billion) accounted for the largest share. Lower but still substantial credits in a range of DM 300-600 million had been granted to Iraq, Poland, Venezuela, Yugoslavia, Chile, Nigeria and Turkey. Even though Deutsche Bank’s exposure was lower than that of other German institutes,^[68] substantial value adjustments were indispensable. By the end of 1983, the bank’s adjustments amounted to DM 2.6 billion. Internal reports listed 50 countries with payment difficulties, for which 30 debt rescheduling agreements involving the IMF, the banks and numerous governments had been initiated in the fall of 1983.^[69] Obviously, this did not mean that all problems had been resolved. In 1984, the Economics Department predicted that the net borrowing requirements of developing countries from banks would increase by about 40 per cent by 1990. This would result in a further adjustment of about DM 1.6 billion for Deutsche Bank, without considering the potential losses of foreign subsidiaries.^[70]

5 International Risk Management and Country Rating

The risks of the Eurodollar markets and Deutsche Bank’s increasing exposure to countries with high sovereign debts and balance of payment deficits were observed with growing concern. How should the bank deal with these risks? Internal discussion about a more elaborate system of risk management began as early as 1977. Since Deutsche Bank was often the leading bank in large, syndicated credits, a thorough assessment of specific country risks related to the banks’ exposure was essential. Until then, information about the specific risks of syndicated lending had usually been gathered by an account manager along with experts from the central credit department. However, most credits had been granted on a case-to-case basis, and often the risk assessment had not been binding. Especially in the case of very large credit consortia, prestige and market strategy had been considered more important than the alleged default risks.^[71] Moreover, Deutsche Bank’s International Department, which was responsible for monitoring country risks, had lacked the statistical and economic expertise to carry out systematic risk assessments. Essentially, it had compiled a country documentation based on eclectically collected data and the travel reports of bank representatives abroad. At this point, a more coherent approach was necessary. The idea of introducing a joint risk assessment scheme under the auspices of the Federal Association of German Banks was dropped, given that German commercial banks were in fierce competition with each other. The management of Deutsche Bank was determined not to give any detailed information on its risk assessment system to other banks.^[72] Moreover, existing statistical data on financial risks provided by the Bundesbank or international agencies such as the World Bank, the International Monetary Fund and the Bank for International Settlements (BIS) often lacked the information required for a more detailed risk survey. The BIS, for instance, only included long-term loans in its statistical accounts, while the commercial banks reported only national data to the BIS. This meant that the exposure of foreign subsidiaries and branches was not considered, although they were highly relevant for the overall exposure of international bank groups.^[73]

Against this background, Deutsche Bank introduced a new comprehensive risk assessment scheme in the autumn of 1977, produced by a joint effort of the bank’s Economics and International Departments.^[74] The new rating system comprised a total of 60 states, but the focus was on 47 developing states including the state-trading countries and the entire Southern European region. These 47 countries accounted for about 40 per cent of the bank’s entire foreign exposure. For each country, the overall risk was measured using a scheme of seven different indicators: the country’s debt service burden, net borrowing requirements, currency reserves, economic growth, foreign trade structure, export expectations and foreign debt. Each country was then assigned to one of five risk categories and re-evaluated annually. Based on the risk categories, limits on lending were introduced for each country. Other important innovations followed. For example, the numerous liabilities were now centrally accounted for, which was technically challenging in view of Deutsche Bank’s complex group structure with its many foreign branches, subsidiaries, and participations.^[75] Foreign exposure included not only loans, but also acceptances, securities and placements which represented claims on residents of a specific country. Centralized accounting allowed Deutsche Bank for the first time to calculate not only specific country risks, but also the bank’s aggregate foreign exposure.

However, the early versions of the new rating system had substantial weaknesses, including that country limits were non-binding for credit decisions. Additionally, it became apparent during the debt crisis that the rating only incompletely captured the real lending risks. For example, a breakdown of the different maturities of liabilities was not possible. Many structural economic and political risks were underestimated even though they had a huge impact in developing countries. For example, countries with a high level of insecurity such as Mexico, Argentina and Brazil had been grouped in the medium risk category (III). Mexico was even upgraded in June 1980 and was in category II until shortly before the outbreak of the debt crisis.^[76] Wilfried Guth, a member of Deutsche Bank’s Board of Directors, stated in October 1982 that the „combination of many negative factors in the case of Mexico had not been foreseen“.^[77]

These negative factors included such external shocks as the second oil price crisis and the impact of the monetary revolution in the USA in 1979. Under its new chair Paul Volcker, the American Federal Reserve Board raised its interest rates to a record of 20 per cent to end inflation.^[78] This shock treatment had global effects and put enormous pressure on countries with high foreign debt and balance of payment deficits. However, the US economy also suffered under the monetary restrictions with rising unemployment, a large number of corporate insolvencies, and credit defaults. For Deutsche Bank, the Volcker shock was sudden and unexpected, and made a reform of existing internal risk assessment schemes indispensable.

In 1982, the Economics Department presented an improved version of the international country rating system. The number of risk groups was increased to seven, and the number of countries under scrutiny to 129.^[79] While industrial countries had until then been excluded from risk monitoring, they could now be monitored as well, even though in practice the focus remained on developing countries. The intervals between the evaluation of the risk assessment were reduced. Instead of one, there were now three evaluations per year. A much more rigorous assessment of risks was now undertaken. The number of risk categories was increased to twelve. Macroeconomic indicators and statistics were more carefully evaluated in making predictions about a country’s potential performance. Finally, political, social, and environmental risks were more carefully considered in the new scoring system. This included a variety of factors such as the quality of the legal system and public institutions, social stability, entrepreneurial culture, climate and environmental problems, the mental openness of the population to foreign influences and a country’s involvement in military, ethnic or religious conflicts. Assessments now addressed not only short-term, but also long-term risks based on a structural analysis of each country’s economic performance. Each overall country score was now subdivided into 12 categories within three groups, measuring the short-term and long-term economic and political risks of each country. The newly elaborated rating scheme led to a reshuffling of risk groups. For example, Brazil, Mexico, and Argentina were now classified in the highest risk group VII, and numerous other countries moved into a more critical category. By 1987, group VII had grown to include 29 countries, making it by far the largest category within the system.^[80]

Fig. 1

Foreign Liabilities of Deutsche Bank per Risk Groups (Mill. DM, 31. Dec. 1987). Source: HADB, ZA18/2: Schultze-Kimmle, Presentation on International Activities, April 19, 1988.

An important innovation of the new rating system was the expanded time horizon. It was intended to consider not only current risks, but also long-term changes in the risk structure of each borrowing country. A country’s past performance – including its record of servicing and repaying external debt – was seen as an important indicator for future default risks.^[81] For this reason, even small changes over the last five years were carefully evaluated to detect possible threats in advance.^[82] Moreover, the bank developed its own phase model which linked changes in the various indicators to specific moments in a country’s debt crisis. This model distinguished eight phases which all countries typically had to pass through during a sovereign debt default, and recommended counter strategies for the bank in order to limit the effects of a credit default on the balance sheet.^[83] In a sense, the static risk analysis that had been used until then was to be turned into a dynamic „early warning system“ to predict debt and payment problems, so that the bank could adjust its lending practice accordingly.^[84]

In addition to producing a detailed quarterly report of all countries’ risks, the Economics Department also drew up more specific statements. These addressed the potential insecurity of a region and the impact of cluster risks or spill-over effects that arose from the default of one country.^[85] In this regard, the economists relied on complex macroeconomic forecasting models, often with a time horizon of six to ten years. For example, a long report drafted in 1984 estimated the expected credit demand of 27 high-risk countries up to the year 1990.^[86] The forecast was linked to projections of key macroeconomic parameters such as interest rates, inflation expectations and a country’s balance of payments. Again, historical data was used for the predictions. The results were used to estimate the potential impact on the bank of a credit default under various macroeconomic scenarios. Moreover, the study supplied the data to calculate Deutsche Bank’s reserve provisions for high-risk countries.

6 Theory vs. Practice: Risk Management and Bank Competition

The statistical analyses and forecasts provided by the Economics Department were not only of academic interest but also served for operational decision-making by the bank’s governing bodies. Alfred Herrhausen, who became the spokesman of the Management Board in 1985, frequently drew on the Economics Department’s studies and often quoted them in international negotiations, public speeches, and interviews.^[87] He also encouraged more systematic planning of all international activities and expanded the bank’s analytical resources. While individual investors, he argued, were able to correct choices and to learn through trial and error, banks required more sophisticated and coordinated strategies of decision-making.^[88]

The new system of risk management was to be the core, designed not only to determine countries’ risks and lending limits, but also to calculate the bank’s overall exposure. The objective was to integrate the banks’ risk management with the structure of the international lending portfolio. Some basic rules were now established for international lending. For example, a single country’s liabilities were limited to 5 per cent of total foreign exposure, while total interest defaults were not to exceed 20 per cent of the previous year’s operating profits. The bank’s total foreign exposure was not to exceed half of its total lending.^[89]

The professionalization of economic risk surveying thus had an increasing impact on lending practices and liability management. Stefano Battilosi has argued that the reorganization of liability and portfolio management was the most important financial innovation that banks introduced during the 1970s and 1980s.^[90] In this process, banks began to analyse the impact of systemic risks on their own balance sheets.

At the same time, risk and liability management became an important strategy in the fierce international competition between commercial banks. Much of this strategy was based on wholesale interbank funding in the Eurodollar market. As this market was still unregulated and thus considered a potential source of considerable balance sheet risks, banks began to monitor their exposure more carefully.

Deutsche Bank was a forerunner in the establishment of new forms of risk and liability management. In 1980, American bankers were already deeply impressed by the sophisticated risk evaluations undertaken by Germany’s leading bank.^[91] With the organizational reform of 1986, Deutsche Bank created two new areas at the management level, responsible for liability structuring and risk supervision.^[92] Nevertheless, the bank was not willing to co-operate with other banking houses or to reveal details of its internal risk monitoring, as this was considered a commercially sensitive advantage. Indeed, the institute was able to consolidate its foreign exposure after 1982 without losing its strong position in international lending. It slowly reduced its unsecured foreign liabilities in high-risk countries, while it increased reserves to cover foreign default risks.^[93] The number of countries for which the bank introduced reserves grew from 2 in 1977 to 47 in 1987 (Figure 2). At the same time, the maturity of aggregate exposure was reduced step by step. By 1987, the bank had shifted a large part of its foreign claims to short-term interbank loans. It was now able to reduce more than 60 per cent of its foreign engagements within three months (Figure 3). Also, the bank had much lower exposure in high-risk regions. Only 25 per cent of its foreign credits were in developing and socialist states (Figure 4).

Fig. 2

Unsecured Lending of Deutsche Bank in Risk Countries (in Mill. DM) and Risk Reserves (in %), 1978-1987. Source: HADB, ZA18/2: Schultze-Kimmle, Presentation on International Activities, April 19, 1988.

Fig. 3

Maturity of Aggregate Exposure of Deutsche Bank (in Bill. DM, 31. Dec. 1987). Source: HADB, ZA18/2: Schultze-Kimmle, Presentation on International Activities, April 19, 1988.

Fig. 4

International Exposure of Deutsche Bank by Region (67.9 Billion DM, 31. Dec. 1987). Source: HADB, ZA18/2: Schultze-Kimmle, Presentation on International Activities, April 19, 1988.

German banks were involved in many international credit negotiations with debtor countries. Since many loans had been granted through large consortia of several hundred banks, the creditors had to coordinate among themselves, which was not easy in view of diverging interests.^[94] For example, some U.S. banks were so heavily involved in Latin America that a quick debt cut would have driven them into insolvency themselves. They therefore tried to delay negotiations to gain time. Banks with lower exposure, on the other hand, were interested in reaching a quick agreement to reorder their balance sheets through a rapid write-down. As already mentioned, this was a strategy of Deutsche Bank, which – unlike many US banks – had no high single country exposure. In addition to the divergent interests of the creditors involved, there were a variety of legal, financial, and political disputes that often led to long and difficult negotiations. To improve coordination, the creditor banks therefore joined together in an informal body – the London Club – which has regularly led negotiations with the debtor states since the 1980s. As a rule, the Club appointed a committee headed by a creditor bank for this purpose. This prestigious task was often performed by large American financial institutions such as Citibank or Bank of America, but also by European institutions including Deutsche Bank.

For Deutsche Bank, this position allowed it to also propose more general solutions to the debt problem. At the annual meeting of the World Bank and IMF in Washington in September 1987, Alfred Herrhausen initiated an international debt cut for developing countries. What appeared to be a generous gesture by Germany’s largest financial institution was in fact a strategic move, given that the bank was far less engaged in Latin America than many US houses.^[95] In any case, it had become unlikely that these countries would ever fully repay their loans. There was no alternative to an overall debt restructuring. Herrhausen suggested that deferred loans should be transformed into debt-to-credit-swaps with guarantees from the World Bank or the IMF – an initiative that was supported by US Treasury Secretary Nicholas Brady and finally accepted by most other commercial banks.^[96]

Due to its provisions to capital against expected losses and a highly diversified foreign exposure, Deutsche Bank managed to survive the crisis relatively unscathed. For example, in the case of Mexico – which in 1989 still had $53 billion in uncovered foreign liabilities – Deutsche Bank had to contribute only a share of DM 170 million out of a total of $7.7 billion that had to be raised by the banks under the Brady Plan.^[97] All in all, the direct financial impact of the debt crisis was less severe for the bank than the uncertainty that paralyzed global capital markets in this period. Until the end of the 1980s, the bank had been able to consolidate its position in international lending. In 1989 alone, Deutsche Bank participated in the issuance of 563 Eurobonds with a value of $108 billion. It also acted as lead manager for many large bonds such as the World Bank’s first global bond issue totalling $1.5 billion.^[98]

7 Conclusion

The case of Deutsche Bank illustrates that establishing an overall risk structure is a key business strategy for banks. Rising market volatility, exogenous shocks and global political uncertainty have provided strong incentives for banks to invest in their information resources and to build up a comprehensive risk management.^[99] As the sovereign debt crisis threatened international commercial banks, diversification of country exposure helped banks to survive. Competition between banks played an important role in this process. Deutsche Bank had been refining its own risk system in the wake of the Herstatt crash and had substantially reduced its foreign exposure in the early years of the sovereign debt crisis. The bank’s strong engagement in Poland and other Eastern European countries triggered a reform of internal risk assessment and lending limits for high-risk countries. This might explain why Deutsche Bank – like other German institutes – had a lower risk profile than many US or British banks which had no exposure in Eastern Europe.^[100]

This case study provides evidence that banks used historical information for their expectation formation. The introduction of new risk assessment schemes was strongly influenced by the experience of rising insecurity in financial markets. Moreover, the evaluation of past repayment records played an important role in determining the lending practice of banks. The case of Deutsche Bank thus confirms older empirical studies of financial memory. They indicate that banks evaluate the credit history not only of individual or corporate clients, but also of states. However, the learning effect from history lasts for only a relatively short time. As we know from empirical studies, the memory of the sovereign debt defaults before the 1930s had only a small impact during the crises of the 1970s and 1980s.^[101] Interestingly, banks not only integrated large shocks or credit defaults into their historical memory, but also systematically monitored incremental changes in past risks. The new indicator system combined the analysis of historical data with an evaluation of a variety of other information sources. The aim was to achieve systematic medium-range forecasts of country risks and foreign exposure.

This paper shows that financial actors rely on both past and present information in forming expectations. What distinguishes them from individual investors is the fact that they constantly review and update their beliefs after obtaining new information. Professional economic forecasting in combination with a comprehensive system of risk and liability management are important features of banks’ expectation formation. The capacity to learn systematically from past errors and to use that experience in their forecasting provides banks with a key advantage over individual investors who often merely extrapolate past developments. Banks – like individual investors – are affected by past experiences, but they usually combine this with more sophisticated forms of informationbased learning.

However, learning from history also has its pitfalls. Two reasons are responsible for this. Firstly, the time horizon of experience-based learning is usually limited to a short period of five to ten years. Recent events are thus often overweighted. Secondly, the instruments of financial institutions to assess risks and form business expectations are usually unchanging once established. Learning from the past is thus based on an existing set of cognitive techniques and models. This might have the effect that new systemic risks are not fully recognized as the established risks schemes are kept in place. Again, Deutsche Bank is a good case in point. The new country exposure scheme established in the 1970s and 1980s concentrated on the Global South and the COMECON area. Except for Italy, no other OECD country was ever included in one of Deutsche Bank’s exposure categories, and, consequently, no credit limits were ever applied to industrial countries. The sovereign debt crises of the 1970s and 1980s would come to have detrimental effects on risk management as well, as the new financial risks emerging throughout the industrialized world from the late 1980s were to be systematically underestimated.^[102] Historical experiences are therefore an ambivalent forecasting instrument as new risks are often not fully addressed.