The possible development of intraday money markets. The analysis in Part II illustrates how banks operating in an RTGS environment may need to adjust the balances on their central bank accounts in the course of the day in order to effect their funds transfers. This adjustment process means that the development of RTGS systems could generate an intraday demand for bank reserves and that positive intraday holdings of settlement balances could have intraday value (apart from their value in meeting any end­of­day requirements). In this respect, the question has been posed whether RTGS systems could contribute to or stimulate the development of intraday money markets. As noted in Section II.2, intraday money markets could serve as an important private sector source of intraday liquidity in the RTGS environment. An embryonic intraday market seems to have emerged only in Switzerland so far in relation to the introduction of an RTGS system (the intraday market in Japan is linked to the periodic net settlement times in BOJ­NET). In the United States, Fedwire has existed as an RTGS system for a number of years, but for a variety of possible reasons, no intraday money market has yet emerged there.

Several factors could determine whether an active intraday money market would develop. With regard to the possible supply of intraday funds, the accumulation of positive balances by banks on their central bank account during the day would create an opportunity for them to lend the funds in the interbank market. Whether banks would prefer to lend funds on a day­to­day or an intraday basis would depend, among other things, on their expected liquidity position during and at the end of the day and on how they assessed the conditions in the money market (e.g. whether they expected the central bank to withdraw reserves from or add reserves to the banking system on that day or whether they anticipated that market factors would significantly affect reserve availability that day).

As regards the possible demand for intraday funds, one factor would be the central bank's policy with respect to the provision of intraday credit. If it offered no such facilities (or if it offered them on unattractive terms), banks would have to rely on interbank markets to obtain liquidity and might have an incentive to borrow in such markets for short periods of time during the day. If banks could obtain intraday credit from the central bank on relatively advantageous terms, this would limit the incentive for banks to engage in intraday money market transactions. This would be the case, for instance, if central bank intraday credit was provided without charge and the opportunity cost of tying up any securities needed as collateral for intraday overdrafts (or for intraday repos) was low.

Another demand factor would be the relative costs of settlement delays in RTGS systems. Such costs could consist of, for example, a loss of customer business (if customers moved their business to banks able to provide a faster service) or the consequences of an inability to settle payments relating to a linked system (such as the payments related to securities settlements). If such costs were felt to be rather low, banks would have little incentive to obtain funds in the interbank market for a limited amount of time during the day or to specify contractual settlement times more precisely. As noted earlier, not all funds transfers in RTGS systems are necessarily time­critical in the sense that they must be settled either at or by a specific point in time or within a limited interval of time during the day. In such circumstances banks could decide to accept a settlement delay and await incoming funds rather than enter into intraday transactions. If, on the other hand, banks were to consider that certain transfers were time­critical, the perceived cost of settlement delays could be higher and this could motivate them to turn to a market for intraday funds. As noted above, for example, settlement of other payment and settlement systems through the RTGS system at specific points during the operating day could be time­critical. Such transfers could thus provide incentives for participants in the RTGS system to enter into intraday funds transactions in order to satisfy liquidity needs within particular periods of time. The (limited) intraday money market in Switzerland is a case in which payments of principal, interest and dividends under a real­time DVP system motivated SIC participants to engage in intraday money transactions to cope with liquidity requirements relating to those payments.

Transaction costs may also be relevant. For instance, it would not necessarily be a cost­effective solution for banks to develop an intraday money market if the transaction costs relating to lending and borrowing funds for periods of several hours or minutes were much higher than the cost associated with available central bank intraday credit. Transaction costs would include the "set­up" costs for the required infrastructure and the development of intraday trading techniques.

Whether an intraday money market would emerge as a market response to the development of RTGS systems is uncertain. The short and tentative analysis provided here suggests that, under current design features of most RTGS systems, the private sector may not necessarily see major advantages in developing "fully­fledged" intraday money markets with characteristics comparable to those of overnight markets. It should be noted, however, that, in principle, the existence of a liquid intraday money market could help banks to operate with lower operating balances and to reduce their reliance on central bank credit facilities. To the extent that intraday money market transactions could be carried out without collateral (or with different collateral), it could also permit banks to economise on collateral needed to secure central bank credit in the form of intraday overdrafts or repos. Intraday money markets could be a private sector solution to facilitating the redistribution of banks' settlement balances during the day in an RTGS system and could therefore contribute to improved efficiency of such systems.

If private interbank markets for intraday funds were to develop, the question for monetary policy would be what relationship such markets might have with established day­to­day money markets in which monetary policy is typically conducted. There are various reasons to expect that in current circumstances any linkages, including those between the interest rates in the two markets, would be weak. It is uncertain whether the level of aggregate reserve balances held during the day and at the end of the day would normally be determined by the same factors. In countries where reserve requirements are in place, for instance, the reserve balances that count towards the reserve requirement are typically measured at the end of the day.

Furthermore, arbitrage opportunities between intraday, traditional overnight and other markets may be limited or constrained by market and settlement practices. For example, technical difficulties in arranging and executing the payment and repayment legs on a precise intraday, overnight, or true 24­hour basis may prevent the establishment of market conventions for contracts that specify precise delivery times for funds. To the extent that markets for non­overlapping periods that span 24­hour periods do not exist, precise arbitrage between very short­term markets may not be possible. Nonetheless, if intraday funding markets do develop alongside existing overnight markets, some rough arbitrage between such short­term markets could be expected, leading to some interrelationships between intraday, overnight and other markets.

Spillover problems. As described in Part II, the central bank could provide banks with intraday credit in support of their RTGS operations in the form of overdraft or intraday repo facilities. The provision of such intraday credit implies that the central bank makes reserves available to the banking system on an intraday basis to accommodate an intraday demand for such working balances. The question is whether such a practice could affect the supply of and demand for overnight central bank credit. In particular, there has often been discussion about whether the terms (interest charge) on which such credit is granted and the amount provided can have an impact on day­to­day money market conditions and are thus relevant for the conduct of monetary policy.

As long as the central bank determines the terms on overnight credit separately and differently from those on intraday credit, conditions in the day­to­day money market should not be directly affected by the terms on which intraday credit is provided. The "segmentation" between the two types of market would remain as long as banks are required to repay all intraday credit at some point in time during the day and they would be unable to substitute overnight borrowings with a combination of intraday borrowings. Nevertheless, there might still be concern about the possibility that intraday credit could be transformed at the end of the day into overnight refinancing. One approach to containing such potential direct spillovers is for the central bank to apply a sanction, for instance in the form of a sufficiently high penalty rate, on any intraday loans or advances that had not been reimbursed by the end of the day. Imposing stiff penalties would not only provide banks with a strong incentive to repay all intraday credit by the end of the day, but would also encourage them to manage their liquidity efficiently during the day. Moreover, it would reduce any possible moral hazard resulting from the provision of standing intraday credit facilities. Such penalty schemes have in fact been adopted or are being envisaged by several G­10 central banks.