The role of the banking sector in the economic process of Lebanon before and after the civil war

Contents

I. Introduction

II. A Political Economy Approach to Finance and Banking
Information-theoretic approaches
The political economy perspective
Hypothesis and Implications

III. Credit Market Development in the pre-War period
Spontaneous Emergence of Banking
The Banking Sector as a Generator of External Rents
Implications for the Process of Capital Accumulation

IV. Endogenous Impediments to Credit Market Development after the War
The Impact of the Civil War on the Banking Sector
Endogenous Rents: The Government enters the Credit Market
The Effects: Crowding Out and Opportunity Costs

V. The Demand Side of the Credit Market
Firms’ Demand for Credit
Other Sources of Finance: Internal Funds & Remittances

VI. Conclusions and Policy Implications

VII. Bibliography

VIII. Appendix

If [Prime Minister] Hariri slipped in the bath and broke his rip there would be a panic until he went on TV to show he was alright.

Middle East Economic Digest, 11 Nov. 1994

I. Introduction

The above phrase captures the current fragility of the link between politics and economics in Lebanon. Accordingly, the economic reform process is stuck in a dead end, being the most precious victim of the current power struggle between Prime Minister Rafik Hariri and President Emile Lahoud. Moreover, Lebanon’s political system, which is a vulnerable calibration of power, distributed along denominational lines, is certainly not very conducive to attaining political and societal consensus. However, despite or maybe because of these conditions, one sector of the economy, namely the banking sector, has done quite well most of the time.

Since independence up to the beginning of the civil war in 1975, the banking sector stood at the heart of Lebanon’s economic strategy – if such a thing ever existed. In any case, it seemed to have worked for a while at least. During the 1950s, growth averaged 6-8 percent and continued on a similar level during the 1960s and early 70s. However, the civil war of 1975-90 obviously put an end to these promising developments and GDP growth was low and often negative. The reconstruction period saw a brief recovery in terms of GDP growth due to large state expenditure until it slowed down in the second half of the 90s and became even negative again in 2000 (see Appendix VIII.15).

What role did Lebanese banks play in this economic process? Did they substantially contribute to the pre-war boom? What is their part in the current slow-down of the economy? The following analysis will shed light on these questions by looking at the links of the banking sector with the real sector during these various periods. To establish these links I will focus on the market for firm-credit. The method I employ is a political economy approach – as opposed to the conventional information theory framework. That is, I will take into consideration the historic evolution of the banking sector and give a stylised account of relevant external and domestic factors.

The next section derives the theoretical framework. Section III looks at developments in banking before the civil war. Section IV analyses the effects of the civil war and subsequent developments. Section V turns to the demand side of the credit market before the paper closes with a conclusion and policy implications in section VI. The Appendix provides the data relating to the graphs in the main text.

II. A Political Economy Approach to Finance and Banking

This section develops the theoretical framework of this essay. First, information theoretic approaches will be briefly tackled. Second, their shortcomings as analytical concepts will lead us to a more holistic approach, i.e. a political economy, or, structural perspective. On these grounds, this essay’s hypothesis is formulated and its implications are sketched.

Information-theoretic approaches

What do economies need financial systems for? The answer to this question seems obvious. Yet, it is not, as indicated by the existence of various theoretical approaches to the issue. Conventional theory, as it can be found in most text books these days, suggests an information-theoretic framework for the analysis of finance and banking (e.g. Mishkin 2004). From this point of view, the financial system simply intermediates between surplus units (depositors/ investors) and deficit units (borrowers/ sellers of equity) by reducing asymmetric information or their effects. Since these two groups have imperfect knowledge about each other, financial markets (including banks and stock markets) facilitate the access to investment or saving opportunities for surplus units and access to funds for deficit units.^[1] The productivity of capital and thus growth is then increased in two ways: First, by collecting information on alternative investment projects. For example, banks investigate the firms to which they lend their funds by examining financial statements. Economy-wide, the result is an improved allocation of funds thus spurring growth. Second, by providing risk-sharing opportunities, investors are induced to invest in riskier but also more innovative technology (Pagano 1993). For instance, by investing into a fund, individuals benefit from the risk-sharing capacity of the fund due to its large portfolio, which may include stocks of blue chip as well as high-risk innovative technology firms. The latter may not have access to funds in the absence of financial intermediaries.

However, one can still observe numerous failures of financial intermediation that challenge academia to provide theoretical explanations. Generally, proponents of the information-theoretic approach employ a principle agent framework to explain why the mechanisms described above do not always operate so smoothly. For instance, Stiglitz and Weiss (1981), by extending Ackerlof’s (1970) concept of adverse selection in used car markets to the realm of finance, argued that even in unrepressed financial markets^[2] credit rationing could occur due to asymmetric information as banks cannot fully observe the credit risk of their borrowers. In order to avoid adverse selection (i.e. advancing loans to an increasing amount of bad borrowers) banks decide to ration credit instead.^[3] Obviously, this leads to a sub-optimal allocation of funds. Similarly, stock markets are analysed from the principal-agent perspective, where information asymmetries between the investor (principal) and the manager (agent) may lead to morally hazardous behaviour, again, resulting in sub-optimal outcomes. A more sophisticated view is put forward by Gertler and Rose (1994). They focus on the idea of a premium for external finance that borrowers have to pay for uncollaterized loans and insurance, due to frictions caused by asymmetric information and problems of enforcement. The higher this premium becomes the more of an impediment it is to productive investment and thus to growth.^[4]

Such information theoretic approaches to financial systems are neither necessarily wrong nor may they contradict the findings of a structural approach that I employ in this essay. However, they are reductionist in the sense that they tend to turn a blind eye to social and political factors and the historic evolution of a particular phenomenon. However, I assume that these are crucial in determining the structure and dynamics of the financial system and thus its impact on – or, more precisely, its interaction with – the real economy. In fact, information theoretic approaches treat the financial system “as an agglomeration of institutions, markets and assets that might or might not exhibit certain informational properties” (Aybar and Lapavitsas 2001: 29). This is a ‘one size fits all’ approach as it does not acknowledge the relationships between the various financial institutions and markets as they grew out of a specific historic development of the country in question.^[5] Policy conclusion based on such an analysis can thus hardly be appropriate.

The political economy perspective

In order to organize our thinking, a few thoughts on the concept of money are useful: In conventional economics money is simply a facilitator of exchange. The analysis starts from the idea of a pure barter economy in which finding a trading partner is quite difficult without the proper means of exchange (e.g. Newlyn 1971: 1).^[6] Money then appears somehow to facilitate this exchange. Yet, not only is this account historically and anthropologically inaccurate, it is also not useful from the perspective of economic analysis because it ignores the existence of credit money.^[7] The Japanese Uno school is based on this concept and attempts to model the general structural features of a financial system with a credit pyramid of four layers (Lapavitsas 2002: 24-26). The analysis starts with the commercial credit market as the first layer, which pertains to the relationship between different commodity producers in the process of capital accumulation. A producer of an intermediate good may accept a deferred payment by the producer of the final product who has to sell it on the market before paying his debt.^[8] By doing so the former spontaneously creates commercial credit for the latter. It is spontaneous in the sense that it endogenously grows out of trade relations - thus it is not imposed from outside, say, by a state agency. Obviously, the amount and structure of commercial credit relations then depend on the specific process of capital accumulation (i.e. size, kind and interaction of firms). Informational asymmetries are not directly relevant here as the advance of commercial credit is based on “relations of trust and commitment arising from regular buying and selling” (Lapavitsas 2001: 37). Commercial credit relations then give rise to financial instruments such as promissory notes or bills of exchange. This leads us to bank credit, which is spontaneously created when, inter alia, bills of exchange are discounted (Itho and Lapavitsas 1999: 86, 92). Thus, the second layer of the pyramid pertains to the relationship between banks and producers or traders. Again, the information-theoretic principal agent perspective is too simplistic as the relations between banks and firms are also determined by the pre-existing networks of commercial credit (as they determine the firms’ demand for banking credit), rather than only by the screening and monitoring of loans through banks. (Lapavitsas 2001: 37). Next, as banks advance loanable money capital to firms they may have to refinance themselves by obtaining money market credit, which represents the third layer of the system. Accordingly, this layer concerns the relationship between banks as loanable capital is traded among banks. The apex of the system is formed by the central bank, which is the “leading bank of the money market” (Lapavitsas 2002: 25). The credit pyramid is complemented by a stock market, which serves the economy’s long term financial needs whereas the nature of the former is generally short-term lending. This concept is illustrated in exhibit 1.

Exhibit 1: The Uno School Credit Pyramid

illustration not visible in this excerpt

Source: My depiction

Hypothesis and Implications

Lebanese commercial banks do not sufficiently fulfil their role as financial intermediaries with respect to the credit requirements of the domestic real sector. From a structural perspective that means that, in Lebanon, both the direct and the indirect links between the second tier of the credit pyramid and the process of capital accumulation are somehow disrupted (see exhibit 1). Flows of loanable capital from banks to firms do not take place to the extent they could and hence the process of capital accumulation – as the main generator of growth and welfare – is largely confined to the commercial credit market in terms of finance.^[9] Assuming a potential for productive investment, the resulting outcome must be sub-optimal. Within this framework, the Lebanese banking sector and its interaction with the process of capital accumulation before and after the civil war shall be analysed in the next three sections. I will explain what causes the above stated disruptions of links between the process of capital accumulation and the credit market, focusing on commercial banks. External developments as well as domestic government policy will be shown to have had decisive influence on the way the commercial banking sector – and the second layer of the credit pyramid in particular – have developed.

III. Credit Market Development in the pre-War period

This section will give an account of the development of the Lebanese banking sector since the end of Ottoman rule to the civil war. Two distinct periods are identifiable: Until the end of the first World War banking activities are characterised by endogenous spontaneity. Subsequently, however, banking sector developments rather become a function of inflowing external rents.

Spontaneous Emergence of Banking

The early history of banking in Lebanon is similar to the one which is to be found in other countries of the Middle East and also in Europe around the eleventh to twelfth century. Given the wide spread of different currencies and far reaching trade relations the profession of the money changer and money lender was vital to the functioning of economic exchange.^[10] These activities naturally focused on local trading centres such as Constantinople, the capital of the Ottoman Empire, Cairo, Mecca and Medina but also on Beirut, which used to be an important trading centre already since Phoenician times (Wilson 1983: 4).^[11] Not surprisingly, money changing and lending was not always clearly separable from merchant activities as traders were often engaged in lending and money changing activities as well. However, over time, specialisation did take place and, moreover, different urban centres produced money exchanging activities, which focused on certain regions. By this token Beirut’s money changers used to deal with coins and precious metals mainly from the Mediterranean region.^[12] Yet, contrary to the historic development in Europe, banks as formal institutions did not appear in Lebanon (or anywhere else in the Middle East) before the early twentieth century. In the words of Udovitch (1979: 255) “we encounter bankers and (…) extensive and ramified banking activities, but we do not encounter banks”.^[13] However, bank-like institutions began to appear in Lebanon with the flourishing of silk trade in the second half of the 19th century, which was then the main pillar of the domestic economy.^[14] Beirut, in particular, served as a mediator between the production of cocoons and silk threads in Greater Syria, and export centres, such as Lyon in France.^[15] That is, producers imported the Mulberry seeds to grow silkworms and exported cocoons and silk threads, mainly to France.^[16] These activities caused the spontaneous emergence of certain banking activities as trade had to be facilitated through, say, documentary credits, which in turn required the establishment of correspondent bank relations abroad. In fact, some of the main silk traders were representatives of foreign banks and sometimes even became their partners. But also production was financed by these quasi-banks as they also engaged in lending activities (Ashi and Ayache 2002: 31). Thus one may conclude that, during this phase, the credit market, as an outgrowth of the silk industry functioned relatively well and linkages with the real sector were solid.

Towards the end of the 19th century, however, the silk industry began to decline due to growing competition from China and Japan on the French market with significantly cheaper artificial silk. Instead of modernizing equipment or moving to other lines of production, production was reduced (Gaspard 2004: 47; Gates 1998: 14). It may thus be of little surprise that after the first World War the above described banking activities had severely declined. According to Badurid-Din (1984: 23), in 1929 only three formal commercial banks were established in Lebanon: one French, one Italian and one Lebanese.^[17] The main player, however, was the Banque de Syrie et du Liban (BSL), which was established in 1920 (at the beginning of the French Mandate) to render monetary services to the governments of Syria and Lebanon. These included the issuance of currency but the BSL was also heavily engaged in commercial banking operations (Badurid-Din 1984: 35).^[18] The crisis in Europe of the 1930s certainly helped the development of the Lebanese banking sector as capital flowed into Lebanon and competition from abroad declined. Moreover, the domestic economy was boosted by military expenditure of the Allied troops present in Lebanon during the Second World War (Gaspard 2004: 52). In the following years, banking activity emerged along a dual structure: Foreign banks dealt with financing foreign trade and related operations, whereas significantly smaller, mostly family owned bank-like businesses and money lenders provided loans to domestic borrowers, especially in rural areas (Gates 1998: 20).^[19] However, the allocation of credit to the latter only benefited rather large landowners, as only they were able to provide the necessary collateral or pay the high interest rates (Sayigh 1982: 71).^[20] It is during the next phase that the domestic credit market declines and lending practices develop a pattern that has not fundamentally changed until today.

The Banking Sector as a Generator of External Rents

After independence in 1943 the banking sector experienced its famous boom, as Lebanon became the financial centre of the region. Although, still in 1945, there were only nine formal commercial banks of which six were foreign owned, in subsequent years, the domestic banking sector expanded dramatically.^[21] By 1967 the number of banks had increased more than tenfold to 93 of which 18 were foreign. Even after the collapse of the Intrabank in 1968^[22] and the subsequent process of consolidation in 1970, there were still 70 banks of which 23 were foreign (see table III.1).

illustration not visible in this excerpt

Similarly, the total amount of deposits with banks increased strongly in the 1950s, 60s and early 70s (see diagram III.1).

[...]

^[1] This approach is derived from the Arrow-Debreu (1954) general equilibrium framework of perfect information: With perfect information, financial institutions are redundant as matching deficit and surplus units know how and where to find each other. Thus allocation is Pareto-optimal without the need for financial intermediation. Yet, if one acknowledges the absence of perfect information, as it is the case in reality, the situation changes and financial intermediaries in the form of banks, insurances, finance companies, investment banks, stock markets, etc. become necessary.

^[2] The concept of ‘financially repressed’ markets was put forward by McKinnon (1973) and Shaw (1973) and refers to financial markets in which interest rates are not driven by market forces but by political directives. Their theoretical work underpinned the Washington Consensus which argued for financial liberalisation as a necessary policy measure to achieve higher growth and development in general.

^[3] Their principle argument is that banks, as they cannot fully observe the credit risk of their borrowers, have to assume that borrowers who are willing to pay high interest rates may be a very risky asset. In fact, the market rate of interest may then correspond with a return for the bank that is below what it could gain if it would ration credit.

^[4] Essential to their argument is the concept of “net worth”. It is defined as the sum of a borrower’s net liquid assets and the collateral value of his or her illiquid assets, including not only tangible assets but also prospective earnings. The higher a borrower’s net worth the lower the premium for external finance because he can offer more collateral value. Obviously, as economic recession are associated with a decrease of individuals’ net worth access to finance during such times becomes even more difficult and may thus spur economic decline even further.

^[5] Instead, the debate centres around the question whether a financial system should be characterized as a bank-based or a market-based system. The latter is often thought to be represented by the financial systems of the UK and USA whereas the latter is said to be found in Japan or Germany (e.g. Demirgüc-Kunt and Levine 1999). In this context Cobham (1995) argues that preference should be given to bank-based systems in developing countries. See Aybar and Lapavitsas (2001) for a critical discussion of this dichotomic perspective.

^[6] This view was introduced by classic economists such as Adam Smith (1999, chap. 4). He argues that trade could not take place between a butcher, brewer and baker unless the very unlikely situation occurs that all of the three exactly require the amount of surplus produce that the other has on offer. Conventional economists today identify three functions of money. These are money as a means of exchange, as a unit of account and as a store of value. However, emphasis is still on the first function (Drake 1980: 44).

^[7] Marxist analysis identifies two types of credit money, namely commercial credit and monetary credit. The former refers to the sale of commodities against promises to pay whereas the latter pertains to the lending of money. See Lapavitsas (1991: 292).

^[8] See Zell (1996) for an extensive discussion of the role of commercial and other credit in pre-industrial England’s woollen industry from a historical perspective.

^[9] Throughout this essay I implicitly subscribe to the Kaldorian view of economics that the industrial and the manufacturing sector in particular, is the main engine of growth. Constraining this sector in terms of credit is thus likely to lead to sub-optimal growth performance. See Thirlwall 2003: 122-23; Pieper 1998:1, 25.

^[10] In Europe as well as in the Middle East, money changers followed the far reaching flows of trade. As traders often bought a commodity in one place but had to sell it in another (were a different coinage was customary) before they were able to pay for it there was a need for credit money which was provided by the invention of bills of exchange. They were accepted across regions and its liquidity was enhanced by the invention of endorsement and discountability. Already in fourteenth century Europe we encounter a highly sophisticated exchange system facilitating trade-relations (Hertz-Eichenrode 2004: 19). Similarly, we find ample evidence for the wide usage of such commercial credit instruments in the Middle East of the eleventh to thirteenth century (see Udovitch 1979).

^[11] It is interesting to note, however, that the profession of the money changer never took root in England as the country was the only which had a single, unified coinage (Kerridge 1998: 1).

^[12] On the contrary, money exchangers in Cairo’s Khan Al-Khalili souk concentrated on coins and jewellery from Ethiopia whereas those in the bazaars of Constantinople handled various European currencies. In the Gulf, Dubai served as the main money changing centre which had specialised in the exchange of the Maria Theresa thalers (which was the currency of the Austro-Hungarian Empire) for the Indian rupee. Interestingly, the former was very common in the region and used in Yemen as the official currency of royalist Yemen until the foundation of the Arab Republic in 1960 (Wilson 1983: 5-6).

^[13] Money changers and money lenders differ from banks in the sense that the former two are part of the informal sector. They are traditionally closely connected to the so-called souk economy where they serve a certain scope of customers. More often than not, customer relations are organized along the lines of ethnic origin or religious affiliation. That is, Beirut’s Maronite Christians would have their own money lenders and so would Sunni and Shia Muslims. Main advantages of this profession are its uncomplicated processing of requests, direct acceptance of jewellery and often more favourable terms due to lower administration costs and the absence of reserve requirements. See Wilson (1983: 8; 14-15).

^[14] It should be pointed out here that the Ottoman Empire was already at the brink of collapse due to severe budgetary problems and an ever devaluating currency. Foreign and private local bank activity was welcomed as the government could not cater the needs of foreign trade. So the government was no obstacle to the development of banking activities in Mount Lebanon (Ashi and Ayashe 2002: 24).

^[15] In 1861 Beirut and its surroundings produced one million kilograms of silk cocoons. This figure raise to 6 million in 1910. Similarly, silk thread production stood at 117,000 kilograms in 1876 and more than quadrupled until 1911 when its production reached 525,000 kilograms (Ashi and Ayashe 2002: 30).

^[16] France was the main absorber of the Lebanese silk exports. Until the outbreak of the first World War it absorbed about 95% of all silk exports (Ashi and Ayashe 2002: 30).

^[17] Today’s Lebanon was then part of French Mandate of ‘Greater Lebanon’ which included Syria. The Mandate was granted to France by the League of Nations in May 1920 and lasted until Independence in November 1943.

^[18] The importance of the BSL in this period is highlighted by the following account by Ashi and Ayashe (2002: 56): “For example, in 1930the total amount of deposits in the bank reached 439 million French francs, a figure that represented 11 times its own capital, twice the amount of money in circulation and around nine times all kinds of credits, including loans granted by the government and guaranteed by the latter.”

^[19] Among the top nine commercial banks with entirely Arab capital in 1930 we find no Lebanese bank. The largest, the Iraqi Rafidain Bank held deposits of US$ 8.6 billion and the smallest, the Kuwaiti Gulf Bank, held US$ 2.1 billion (Sayigh 1982: 71).

^[20] There was a half-hearted attempt by the government to supply credit to the agricultural sector. However, practices were often inefficient. For example, state-guaranteed credit given to landlords who then advanced the very same capital to peasants at far higher rates of interest. Few of the latter could afford them in the first place (Gates 1998: 22).

^[21] However, as there was virtually no regulatory framework during the French Mandate bank-like institutions had existed already then to a quite considerable extend. Close to the end of the Mandate in 1940 there were 36 local quasi-banks and 89 cooperative institutions (Ashi and Ayache 2002: 62).

^[22] The Intra Bank was then one of the largest local Banks. Its collapse followed the classic pattern of a bank run, which plunged the bank into a liquidity crisis due to a severe maturity mismatch. That is, the bank had mainly lent long-term against primarily short term deposits on the liability side. Although the crisis was prevented from spreading to the sector as a whole, it triggered a series of new banking legislation which generally tightened the rules. See Wilson (1983: 61) and Ashi and Ayache 2002: 123-126.