The banking and payment system in centrally-planned economies was good at what it was designed to do - namely to closely monitor enterprise performance relative to the financial plan and, in that capacity, provide cash for personal transactions and payment orders for inter-enterprise payments. It is thus not surprising that, when the goals are altered - as they are in a transitional economy, a number of problems arise. Six important issues needing to be addressed are outlined in Table 4.

Structure. The shift from a state-owned mono-banking structure, to one where there is more than one bank and each is privately owned, means that the legal structure underlying both the banking system and the payment system has to be considerably expanded for non-cash payments. This involves property rights (for private ownership), bankruptcy law (because banks can fail), and a legal/regulatory structure which clearly spells out the rights and liabilities of payors and payees in the payments chain. In this process, it is important that these laws and regulations permit new and cost-effective payment instruments to develop and evolve if they meet better the payment needs of households and enterprises in a market economy. The old procedure, where the government decides, in isolation, what payment arrangements should be permitted and then tightly controls all aspects of a payment cycle, will not achieve the desired results: household, enterprise, and banking needs should be given a large weight when the payment system is being restructured. This is more important for inter-enterprise and interbank payments, which typically rely on non-cash instruments, than it is for households which will generally rely on cash for most of their personal transactions.

Purpose. Transitional economies recognize that there is a time value to money. This recognition, and the fact that the payment system is no longer an enforcement control device for the financial plan, means that the payment system will have to be adjusted to serve two purposes: the accounting purpose of providing an audit trail for inter-enterprise and other transactions plus the timely value transfer purpose to minimize the enterprise and household float costs associated with transactions.

The costs of untimely payments can be substantial. In Russia, it has been estimated that upwards of one-third of the increase in the money supply has been absorbed and used only to clear payments. Because of credit risk and fraud, the PDO (a debit transfer instrument initiated by the payee (see Figure 1) has been outlawed. In its place is the payment order - a credit transfer - initiated by the payor. While in the past a payment transaction was usually completed in a matter of days, currently in Russia weeks can elapse from the time when an enterprise account at a bank is first debited to when the credit finally enters another enterprise's account and can be used. This time gap is the result of increased lags in processing, collecting, and settling payments within the banking system as well as increased delays by enterprises in making requested payments (evidenced by rising inter-enterprise arrears).

The cost to enterprises of delayed payments is either: (a) the cost of borrowing working capital to cover the time needed to collect and settle payments which, if received earlier, could have been used as working capital; or (b) if working capital reserves are sufficient, the lost interest revenue which would have been earned on collected funds if these funds were collected earlier. In addition, the cost of (b) is increased, while the cost of (a) is reduced, when interest rates do not fully reflect inflation and the reduction in purchasing power which results. When payment transfers are speeded up, these enterprise costs are reduced. However, if the proportion of the money supply tied up in the payment system falls, the proportion in the hands of the public rises. This change is equivalent to a rise in the money supply and needs to be coordinated with the monetary authority.

Enterprise payments. Paper-based enterprise payments may be adequate when values are small but are expensive when values are large. In the short-run, the timeliness of paper-based enterprise payments can be greatly improved through the development of machine-readable MICR encoding for checks or payment orders and processing these items using high speed reader-sorters. Collection times can be markedly reduced by switching from the post office to dedicated motor couriers, domestic commercial airlines, and even dedicated air couriers. In the longer-run, particularly within and between major domestic trading centers, it will be cost-effective to develop a personal computer or mainframe-based telecommunications infrastructure that will accommodate enterprise, money market, and international large value payments on a same-day basis.

Guarantees. As payments are no longer from one state-owned enterprise to another, the state no longer guarantees either the viability of the banking system nor the credit-worthiness of enterprise payments. Private ownership comes at the cost of a loss of state guarantees for the payment system. Thus receivers of payments need to develop procedures to assess the credit risk of the counter-party they deal with in transactions. This is not a simple process, even in a market economy: it requires information on a counter-party's credit history, its financial viability, and a thorough understanding of the rights and liabilities of all parties in different types of payment transactions. When the legal underpinnings of the payment system is itself uncertain, as it can be in a transitional economy, credit risk is even more difficult to judge. In this event, it is quite important to have personal knowledge of the counter-party and his credit history prior to engaging in trade.

Settlement. With the breakup of a mono-banking structure into numerous privately-owned banks, enterprises may have more than just one account and deal with more than just one bank. In market economies, the central bank provides for inter-bank settlement of large value payments using either real-time gross settlement (RTGS), where each transaction is settled when sent over a large value network, or end-of-day net settlement, where only the multilateral net position of an entire day's series of large value transactions is settled. To reduce systemic risk, net settlement networks limit the net debit exposures of bank participants using real-time net debit caps, develop loss-sharing agreements allocating the cost of a settlement failure, and/or require the posting of liquid collateral to cover the single largest possible net debit in a settlement failure. These and other payment risk-reduction procedures have been developed by G-10 central banks under the auspices of the Bank for International Settlements (referred to as Lamfalussy standards).

On a RTGS network, adequate balances are held to clear all payments at the time they occur. However, the opportunity cost of holding idle payment clearing balances exceeds the expense of positing liquid collateral (which can earn an interest return). Thus net settlement of large value payments backed by liquid collateral is a lower cost alternative to RTGS and can be structured to be just as free of systemic risk. But systemic risk is only one problem. As noted above, same-day (or even next-day) payment and settlement of enterprise large value transactions should be an important goal to reduce payment float.

Retail payments. Retail payments can be made more convenient and safer if many household payments that previously relied almost entirely on cash were expanded to permit check or GIRO payments and perhaps even electronic point of sale debit or credit cards. While cash will likely remain the dominant payment instrument in terms of the number of transactions (e.g., Table 1), larger value household transactions and transactions where face-to-face contact is either inconvenient or unnecessary represent areas where alternative payment instrument use is likely to arise.

An important side effect of a reduced use of cash for larger value household payments is a corresponding reduction in the future growth of seigniorage benefits to the government. When the government issues currency, it gives itself a long-term "loan": the newly issued currency is used to purchase real goods and services for the government (the value of which far exceeds the actual cost of printing the currency) and the public may never redeem the currency (although there will be recurring costs of currency replacement over time). While the stock of currency in circulation need not fall as households develop and use alternative non-cash payment instruments, the rate of growth of currency - and hence the benefit from seigniorage - should slow over time. As seigniorage creates revenue for the government, alternative sources of revenue must be developed. An expansion of central bank credits to government enterprises is not the answer. Although rapid growth in either cash or central bank credits can create an inflation "tax" on the economy (since the real value of debt is reduced), cash represents a "loan" from the public to the government that is unlikely ever to be repaid (seigniorage) while central bank credits to government enterprises represents a loan that if not repaid, reduces government purchasing power in the future.