# Organisational Structures - Modern Banking

The intermediary and payments functions explain why banks exist, but another question to be addressed is why a bank exhibits the organisational structure it does. Profit-maximising banks have the same objective as any other firm; so this question is best answered by drawing on traditional models. Coase (1937), in his classic analysis, argued that the firm acted as an alternative to market transactions, as a way of organising economic activity, because some procedures are more efficiently organised by ‘‘command’’ (e.g., assigning tasks to workers and coordinating the work) rather than depending on a market price. In these situations, it is more profitable to use a firm structure than to rely on market forces.

The existence of the ‘‘traditional’’ bank, which intermediates between borrower and lender, and offers a payments service to its customers, fits in well with the Coase theory.

The core functions of a bank are more efficiently carried out by a command organizational structure, because loans and deposits are internal to a bank. Such a structure is also efficient if banks are participating in organised markets. These ideas were developed and extended by Alchian and Demsetz (1972), who emphasised the monitoring role of the firm and its creation of incentive structures. Williamson (1981) argued that under conditions of uncertainty, a firm could economise on the costs of outside contracts.

Banks and the Principal Agent Problem

The nature of banking is such that it suffers from agency problems. The principal agent theory can be applied to explain the nature of contracts between:

• the shareholders of a bank (principal) and its management (agent);
• the bank (principal) and its officers (agent);
• the bank (principal) and its debtors (agent); and
• the depositors (principal) and the bank (agent).

Incentive problems arise because the principal cannot observe and/or have perfect information about the agent’s actions. For example, bank shareholders cannot oversee every management decision; nor can depositors be expected to monitor the activities of the bank. Bank management can plead bad luck when outcomes are poor.

Asymmetric information, or differences in information held by principal and agent, is the reason why banks face the problem of adverse selection because the bank, the principal, normally has less information about the probability of default on a loan than the firm or individual, the agent. Though not shown in Figure below, the presence of adverse selection may mean the supply of loans curve is discontinuous at some point. Adverse selection is the reason why the supply curve is discontinuous or even backward-bending (with respect to certain borrowers), and shows that bankers are more reluctant to supply loans at very high rates because as interest rates rise, a greater proportion of riskier borrowers apply for loans. The problem of adverse incentives (higher interest rates encouraging borrowers to undertake riskier activities) is another reason why banks will reduce the size of a loan or even refuse loans to some individuals or firms.

Moral hazard is another problem if the principal, a customer, deposits money in the agent, a bank. Moral hazard arises whenever, as a result of entering into a contract, the incentives of the two parties change, such that the riskiness of the contract is altered. Depositors may not monitor bank activities closely enough for several reasons. First, a depositor’s cost of monitoring the bank becomes very small, the larger and more diversified is the portfolio of loans. Though there will always be loan losses, the pooling of loans will mean that the variability of losses approaches zero. Second, deposit insurance schemes reduce depositors’ incentives to monitor the bank. If a bank can be reasonably certain that a depositor either cannot or chooses not to monitor the bank’s activities once the deposit is made, then the nature of the contract is altered and the bank may undertake to invest in more risky assets than it would in the presence of close monitoring.

Shareholders do have an incentive to monitor the bank’s behaviour, to ensure an acceptable rate of return on the investment. Depositors may benefit from this monitoring. However, even shareholders face agency problems if managers maximise their own utility functions, causing managerial behaviour to be at odds with shareholder interest. There are many cases of bank managers boosting lending to increase bank size (measured by assets) because of the positive correlation between firm size and executive compensation. These actions are not in the interests of shareholders if growth is at the expense of profitability.

Relationship Banking

Relationship bankingcan help to minimise principal agent and adverse selection problems. Lender and borrower are said to have a relational contractif there is an understanding between both parties that it is likely to be some time before certain characteristics related to the contract can be observed. Over an extended period of time, the customer relies on the bank to supply financial services. The bank depends on long-standing borrowers to repay their loans and to purchase related financial services. A relational contract improves information flows between the parties and allows lenders to gain specific knowledge about the borrower. It also allows for flexibility of response should there be any unforeseen events. However, there is more scope for borrower opportunism in a relational contract because of the information advantage the borrower normally has.

The J¨urgen Schneider/Deutsche Bank case is a good example of how relationship banking can go wrong. Mr Schneider, a property developer, was a long-standing corporate client of Deutsche Bank. Both parties profited from an excellent relationship over a long period of time. However, when the business empire began to get into trouble, Schneider was able to disguise ever-increasing large debts in his corporation because of the good record and long relationship he had with the bank. Schneider forged loan applications and other documents to dupe Deutsche and other banks into agreeing additional loans. In 1995, he fled Germany just as the bank discovered the large-scale fraud to cover up what was essentially a bankrupt corporation. After nearly 3 years in a Florida prison, Mr Schneider gave up the fight against extradition and was returned to Germany to face the biggest corporate fraud trial since the end of the Second World War. In 1998, he was convicted of fraud/forgery and given a prison term of 6 years, 9 months. The judge criticised German banks for reckless lending. Outstanding loans amounted to $137 million. Deutsche Bank apologised for improper credit assessment, especially its failure to follow proper procedures for loan verification. Transactional or Contract Banking An arms-length transactionalor classical contractis at the other extreme and gives rise to transactional banking – where many banks compete for the customer’s business and the customer shops around between several banks to find the best deal. Little in the way of a relationship exists between the two parties – both sides stick to the terms of the contract. A transactional contract deters opportunistic behaviour and because each contract is negotiated, both parties can bargain over terms. On the other hand, information flows will be significantly curtailed and the detailed nature of the contract reduces the scope for flexibility. It is important to treat the definitions given above as two extremes, at either end of a spectrum. In reality, most banks will offer a version of relationship banking to some customers or apply it to some products, while contract-like banking is more appropriate for other clients and/or services. For example, virtually all customers who enter into a loan agreement with a bank will sign a legally binding contract, but if the customer has a good relationship with the manager and a good credit history, the manager is likely to allow a certain degree of flexibility when it comes to enforcing the terms of the contract. For new clients, the manager will be more rigid. Relationship banking is most evident in countries such as Japan and Germany, where there are cross-shareholdings between banks and non-financial corporations. In other countries, including the USA and the UK, classical contracts are the norm. In Japan and Germany, the close bank–corporate relationships were, in the 1970s and 1980s, praised as one of the key reasons for the success of these economies. However, in the 1990s, relationship banking declined because of global reforms, which increased the methods for raising corporate finance and the number of players in the market. Furthermore, the serious problems in the Japanese financial sector that began in 1990 have undermined keiretsu, the close relationship enjoyed by groups of firms, including a bank. The bank plays a pivotal role in the group because it provides long-term credit to the main firm and its network of suppliers, as well as being a major shareholder. The bank also gives the keiretsu advice and assistance in overseas ventures. With the steady rise in the number of key banks facing bankruptcy, primarily as a result of problem loan portfolios, and a drastic reduction in the market value of banks’ equity portfolios due to the prolonged decline in the stock market, the relationships between banks and corporations have been seriously undermined. Payment Systems: A Byproduct of the IntermediaryProcess One theme of this is that banks differ from other financial firms because they act as intermediaries and provide liquidity. Banks require a system for processing the debits and credits arising from these banking transactions. The payment system is a byproduct of intermediation, and facilitates the transfer of ownership claims in the financial sector. Credits and debits are transferred between the relevant parties. In the UK alone, there were over 28 billion cash payments in 2001, but they are expected to decline to 24 billion by 2010. £113 billion was withdrawn from the 34 300 Automatic Teller Machines (ATMs) in 2000.In the same year, there were 3 billion plastic card transactions with UK merchants. However, there are two key risks associated with any payment. Banks must manage the following. • Liquidity risk: The settlement is not made at the expected time so that assets/liabilities cannot be transferred from one agent to another via the system. • Operational risk: Arising from the threat of operational breakdowns, preventing timely settlement. For example, the hardware or software supporting the system may fail. System breakdowns can create liquidity risk. Given the open-ended nature of the term, it is difficult to provide a precise definition, which makes measurement problematic. In the UK, payments are organised through the following. • APACS (Association for Payments Clearing Services): An umbrella organisation formed in late 1984, and made up of BACS, CCCL and CHAPS. It was supposed to allow relatively easy entry of banks into the UK payments system. Membership is offered to all participants with at least 5% of total UK clearing. Financial firms that do not qualify for membership but offer products requiring clearing and payments are made associate members. • BACS Limited: An automated clearing house for non-paper-based bulk clearing, that is, standing orders, direct debits and direct credits. Fourteen direct members sponsor about 60 000 other institutions to use the system. As can be seen from Table, BACS clearing volumes stood at 3.7 billion in 2002. • CCCL (Cheque and Credit Clearing Company Limited): Responsible for paper-based clearing, i.e., cheques. In 2002, there were 2.4 billion cheque transactions (see Table below), which is forecast to fall to 800 million by 2012.10 • CHAPS: Provides Real Time Gross Settlement (RTGS) for high value payments, and is the second most active in the world. In 1998, the average value of transactions processed was £2.3 million, compared to £552 for BACS. In 2000, there were some 25 million • UK: total transactions by volume (millions) Euro was formally launched in January 2001, to process euro payments between members, with monthly volumes of 280 000, valued at 3600 million euros. It also provides the UK link to TARGET (see below). The real time nature of the settlement eliminates settlement/liquidity risk, unlike BACS, which settles payments in bulk. • CLS: Created to reduce risks associated with payments involving another currency. It will gradually replace the standard foreign exchange settlement method, where a correspondent bank is used. In 2002, CLS introduced real time payment for foreign exchange transactions. • CREST: Settlement of Securities. Central bank-related transactions moved to real time in 2001, and the idea is to introduce it for all money market instruments – payments are still made at the end of the day on a net settlement basis. The London Clearing House (LCH) acts as a central counterparty for transactions on the financial exchanges, and for some over the counter markets. At the end of 2003, LCH merged with its Paris counterpart Clearnet, creating Europe’s largest central counterparty clearing house. It will go some way to creating a pan-European clearing house, reducing the cost of cross-border trading in Europe. Use of Cards and ATMs In the mid to late 1990s, there was a continued rapid growth in the use of cards instead of cheques. This point is illustrated in Table above. This table also illustrates that cash payments over the decade and into the new century are fairly stable, and ATM withdrawals have more than doubled. Cash payments remain the dominant payment method, making up three-quarters of all payments, and their dominance will continue, though there might be a slight decline once social security benefits are paid directly into accounts. The use of cheques as a form of payment has fallen dramatically, as households and businesses switch to the use of plastic cards or direct debit/credit. About 3% of card transactions were via the internet in 2002, and by 2012, APACs is forecasting this to grow to 10%. The ATM network in the UK is run by LINK, which is jointly owned by the banks and building societies. Via LINK, customers have access to over 34 000 ATMs. There are two credit card schemes: Mastercard, owned by Europay, and Visa, part of Visa International. There are also two debit card schemes: Switch and VisaDebit. Cruickshank (2000) reports that the payment schemes (APACS, Visa, etc.) and ATM network are dominated by the ‘‘big four’’ banks because the size of shareholdings is normally determined by the volume of transactions in a given scheme. Cruickshank criticised the consequences of this control, which was to take advantage of their monopoly position. Other users of the network were being charged excessive amounts, which had to be passed on to their customers or absorbed in their costs. For example, internet banks had to pay twice as much for access to the system as the big four, and retail outlets were charged excessive prices to offer a direct debit/credit card service to their customers. Cruickshank reported that the fee charged bore no relation to the cost of the investment undertaken by the big four. The big four banks paid the lowest prices to use the system, and, for a brief period, account holders faced charges if they used a rival’s machine, though a vociferous public campaign forced banks to largely abandon this practice. Cruickshank recommended the establishment of an independent regulator for the payment systems: Paycom. Access would be via a licence, the price of which would reflect the cost of use by a given bank. It could also ensure entrants were financially sound, to minimise settlement and liquidity risk. For example, with the exception of CHAPS, the systems are not based on real time gross settlement, so any bank that failed while it was still using the payments system could strain the liquidity of the system. The British government accepted the need for reform, and referred the matter to the Office of Fair Trading. It has announced the introduction of PaySys, a rule-based system to regulate the payments industry (the Treasury will draft the relevant details), which does not go as far as the ‘‘public utility’’ approach represented by Paycom. An alternative is the ‘‘competing network’’ model, whereby there are several large networks that compete for banks to join them. The clearing system in the United States is quite different. The Federal Reserve Bank operates a number of cheque clearing centres, which are responsible for about 35% of US cheque clearing, which amounted to$13.4 billion in 1998.16 Private centre arrangements made between banks account for another 35%, and about 30% is cleared by individual banks. In 1998, $16 billion worth of electronic payments were processed through one of 33 automatic clearing houses (ACHs) run by the Federal Reserve or one of the private ACHs. International interbank transactions are handled by CHIPS, the Clearing House Interbank Payments System. It is run by the privately owned New York Clearing House Association. CHIPS uses multilateral netting. Until 2001, all net obligations were cleared at the end of the day, but a new bilateral and multilateral algorithm means most payments will be settled promptly through a given day, thereby reducing settlement risk. In 1998, there were roughly 60 million settlements, with a total value of about$350 trillion.

Fedwire is operated by the Federal Reserve and allows banks (that keep deposits or have a clearing facility with the Federal Reserve) to send and receive payments. With more than 11 000 users (1998) there were over 98.1 million transactions worth $328.7 trillion. Fedwire has offered net settlement facilities since 1999, which has reduced members’ exposure to settlement risk. In Europe, TARGET (Trans-European Automated Real Time Gross Settlement Express Transfer System) was set up in response to the European Monetary Union. It means central banks can transfer money within each EU state. It consists of 15 national RTGS systems, the European Central Bank Payment Mechanism (EPM) and SWIFT, which interconnects these systems. Since the settlement is immediate, in real time, it eliminates settlement risk, because the payments are deducted from and credited to the relevant accounts immediately. TARGET is viewed as a harmonised system, and greater harmonisation is expected in the future. According to BIS (2003f), TARGET processes over 211 000 payments each day, valued at ¤1.3 trillion. Though TARGET eliminates settlement risk, operational risk is considerable. For example, in 1999, a system error at one of the very large banks meant it was unable to process payment orders for foreign exchange, money market transactions, securities settlement and customer payment. The backup system also broke down because it relied on the same software. Manual systems could not cope, so that many large value payment and securities orders were not settled until the next day – this operational breakdown effectively recreated settlement risk. Apart from the TARGET arrangement for central banks, the situation in Europe looks bleak. With the introduction of the euro in 2002, there is a need for a payments system that allows for quick settlement within Euroland. Instead, there is a plethora of bilateral agreements between different banks. Eurogiro was set up in 1992 by 14 countries’ giro clearing organisations, and a similar system, Eufiserv, operates among the European savings banks. Some moves have been made to link CHAPS with its equivalent in France (SIT), Switzerland (SIS) and Germany (EAF), but no formal agreement has been reached. The large number of independent arrangements (that do not include all banks) will hamper cross-border settlement even if banks are all using one currency, the euro. The cost of cross-state settlement in Europe is estimated to be substantially higher than in the United States. Increasingly, the responsibility for payments and securities clearing is being unbundled from the traditional bank functions, and given to a third entity, which is not necessarily another bank. These firms are providing a service to banks: processing settlements and securities for a large number of banks, reducing banks’ back office operations. In other words, back office functions are becoming the sole activity of certain firms, which the banks pay, rather than having their own back office operations. According to BIS (2001, p. 310), in the USA, the top five non-bank service providers make up 20% of the outsourcing market. An International Comparison of Payments Technology Figures below illustrate how the pace and form of payments-related technological innovation has varied widely among the different industrialised countries. Figure below shows that ATMs are more plentiful in Japan and North America than in Western Europe. In Europe, Denmark has the fewest ATMs relative to population, followed by the UK and the Netherlands. The other European countries are roughly the same. The change in the UK is surprising because, in the 1980s, it was one of the leading ATM countries in Europe. It is consistent with the large number of branch closures in the UK, and ATMs have not spread in sufficient numbers to other sites, such as supermarkets, rail and petrol stations. Turning to Figure below, Germany stands out as having relatively few Electronic Funds Transfer at Point of Sale (EFTPOS) machines, followed by Italy, the USA and Portugal. However, while the ratio of population to EFTPOS is 466 in Germany, it is half that in the USA. Countries with relatively more machines include Spain, Switzerland, Canada and France. Switzerland, Japan and the USA have relatively high paperless credit transfers (Figure below), while some of the continental European countries rank at the bottom – France, Portugal, Italy and Belgium. Figure below shows the USA, Canada and the UK have the highest value of payments by credit and debit cards, with some of the continental countries lagging Average population per ATM. Average population per EFTPOS machine. Ratio of value of paperless credit transfers to nominal GDP. behind – especially Germany, Italy and Spain. The use of credit and debit cards in Japan is also low, compared to other countries. Correspondent banking and custody services are also part of the payments system. Correspondent banking is an arrangement whereby one bank provides payment and other services to another bank. Reciprocal accounts, which normally have a credit line, are used to facilitate Ratio of value of payments by debit and credit cards to nominal GDP. payments through the correspondent bank. Custody services involve the safekeeping and administration of securities and other instruments on behalf of other banks or customers. Globally, the number of banks offering these services has declined, as a small number of large banks dominate an increasingly consolidated market. For example, the Bank of New York has opted to be a niche player, offering global custody services to other banks, managing$6.3 trillion worth of custody assets in 2000. Banks specialising in these services normally have sound reputations, offer a fairly large range of products and services that are easily obtainable, participate in key payment and settlement systems, and can raise liquidity.