 |
|
by Jerry L. Jordan and Edward J. Stevens
Federal Reserve Bank of Cleveland
Innovation is normal when it comes to the subject of money. For centuries, the fundamental forces of technological change and market competition have been changing both the forms in which money is held and the methods by which its ownership is transferred. Looking ahead over the next century, there is no reason to expect anything different. It is not possible to predict how fast things will change, or exactly what forms monetary innovations will take. Rather than predict, we want to share some of our uncertainty about the implications for the central bank of some potential changes in monetary mechanisms. Three points:
- Past monetary innovations have tended to increase the income velocity of central bank money. Potential innovations being discussed at this conference will continue this trend in the 21st century.
- The income velocity of central bank money might approach infinity in the next century, as holdings of central bank currency and deposits shrink in importance --just as did holdings of commodity money in the 20th century.
- How the declining demand for central bank money might influence the role of the dollar as a standard of value is not yet known. We are encouraged, however, that this is a topic on which both theoretical and empirical economic research is focusing.
After a brief overview of innovations in money regimes, we elaborate on each of these points.
Innovations in Money Regimes
Monetary history records repeated innovations in the assets that have been readily transferable stores of value and in the mechanisms used for transferring those asset values. As recently as the 19th century, “money” meant both the money storage asset and the money transfer mechanism. Full-bodied commodity money, fractional coin, government fiat currency, and bank notes all provided joint products: assets for storing value and, at the same time, vehicles for instantaneous, face-to-face transfers of value, with finality.
The inconvenience of face-to-face payments in an increasingly integrated national economy was avoided by accepting the cost and risk of delayed payment finality. Local clearinghouses became part of the transfer technology, facilitating both the clearing and settlement of checks drawn on asset values stored in local bank deposits. At greater distances, payors could purchase “exchange,” consisting of a local bank's check drawn upon its distant correspondent, which could then be mailed to the payee, bringing the Post Office into the transfer mechanism. The United States Post Office, as well as some commercial enterprises like American Express, operated independent money order services for transferring money values over distances.
In the first decades of this century, telegraphic transfers of balances for same day value were the cutting edge of money technology. The dominant retail money technology was still shifting to paper checks drawn on commercial bank demand deposits. Over the second half of this century, rapidly declining costs of computing and telecommunications increasingly made it feasible for a wider variety of assets to be exchanged very quickly, fulfilling some of the “store of value” functions of money, but, only indirectly capable of being transferred to third parties.
Successively broader definitions of money (M2) have recorded the widening field of effectively monetized assets. As we come to the end of the century, M2 includes NOW [1] account and money market deposit account balances at banks and nonbank depository institutions, and shares held in money market mutual funds, in addition to the original combination of currency, demand deposits, small time, and savings deposits. This certainly commingles “dollars” and assets denominated in dollars. Moreover, all these assets can be transferred to third parties by paper or electronic payment orders directly, or, at least, so rapidly from one kind of account to another as to be indistinguishable from a direct third-party transfer.
The increasing speed with which transactions can be effected has been part of the innovation process in the money industry. At the cutting edge of money technology, corporate America is moving beyond batch processing and air couriers, to networks for integrated accounting and payments processing systems. Standing at the entrance to the 21st century, the definition of an instantaneous money transfer--not by check, but by ATM or direct computer connection--is moving inexorably toward "real time," on a par with exchanges of currency, but without the need to be physically face to face. And the closer technology brings us to real-time payments, the closer we are to genuine 24-hour banking and trading, and a worldwide set of assets that might be used for wealth storage, at least for those who are willing to accept some currency risk.
The proliferation of money assets and increasing speed of money transfers are two trends that clearly will persist into the next century. So, too, will the third trend--the elimination of regulatory and other legal restrictions on the money industry erected by governments. On a global scale, modern communications technology assured the free flow of information through the Iron Curtain and flattened the Berlin Wall. In the United States, that same technology has flattened artificial walls between groups of depository and other regulated financial institutions, and between regulated and unregulated institutions. Telecommunications-based information technology made it cheaper and cheaper to avoid costly regulations. Initially, this perpetuated a kind of cat-and-mouse game between regulators and markets. In banking, at least, that game now is ending. For example, last year brought the advent of accounting programs that sweep reservable deposits temporarily into nonreservable form for all the retail deposits of a bank, not just its corporate cash management customers. This change is reducing, and may finally eliminate, an effective reserve requirement tax.
The rapid spread and ultimate success of sweep programs epitomizes 50 years of experience with erecting and flattening arbitrary regulatory walls between industries. In the long run, regulation doesn't work. Regulation created profit incentives for banks to avoid reserve requirements, deposit rate ceilings, and line of business restrictions by taking their money business outside the traditional orbit of the banking industry, lest it be taken there by nonbanks. The same restrictions created incentives for the competitors of banking to bring the business of banking into the orbit of nonbanking industries. Sometimes the banks prevailed; sometimes nonbanks prevailed. Never did the regulators prevail, and the walls have come tumbling down. Regardless whether Congress ever removes Glass-Steagall restrictions, the long-run futility of using regulations to enforce arbitrary restrictions seems well documented.
Past Innovations and the Income Velocity of Central Bank Money
Prospects for electronic moneys seem reasonably good, at least initially because they might build on foundations already laid by a firm's existing, unique product line. A subway system might move from a stored value fare card toward a more general purpose stored value card. A long distance carrier might build on its nationwide commercial and consumer network.
The predicted impact of such innovations on the central bank has a familiar ring. Around the turn of the century, as the use of checking accounts became widespread, monetary analysts recognized that these deposits were substitutes for traditional gold and paper money. To account for the effect of this substitution on what we now call monetary policy, discussion focused on the resulting increase in the income velocity of what was the equivalent of today's central bank money. Then, in the 1950s, the thrift industry enjoyed overwhelming competitive success in providing assets that were so liquid as to be good substitutes for checking account deposits. Discussions of the impact of this substitution recognized that the income velocity of central bank money increased when thrifts issued monetary assets without holding significant reserves of central bank money. Then, again, in the 1970s, discussion focused on electronic funds transfer systems (EFTS). This time the concern was about increases in the velocity of money resulting from reduced inventories of money. Electronics was expected to allow existing moneys to be transferred with greater speed and precision over emerging telecommunications networks that would link merchants and customers and banks.
Innovations in monetary assets and transfer systems produce short-run variations in the income velocity of central bank money, which tend to complicate monetary policy decisionmaking. So, too, might increases in velocity over the long run. Monetary targets are more difficult to define and achieve during the transition from one type of monetary regime to another. Higher velocity might make it more difficult to maintain financial stability, unless there were comparable increases in the precision with which the central bank could control the supply of its monetary liabilities. We are not aware of anyone who suggests, however, that a long-run increase in the income velocity of the monetary base has led the Federal Reserve seriously astray, or allowed the purchasing power of the dollar to fluctuate as much as it has in the 20th century, or to decline as much as it has in the postwar period. However rough, the policy process can still operate through feedback directly from movements in the observed price level.
Checking the facts against the expectation of increased velocity is instructive. Despite all the new types of monetary assets and transfers, base velocity has increased by only about two-thirds, or at a compound average annual rate of less than 1.5 percent since 1959, when the current data series began. Velocity of the deposit component of the base more than quadrupled, but that of the currency component barely changed. These results are distorted, however, by outflows of U.S. currency to foreign markets that have needed a more reliable money than provided by their own monetary authorities. [2] If we illustrate our suspicion by assuming that the percentage of currency held abroad has gone from a negligible amount in 1959 to almost two-thirds today, then the income velocity of domestically held central bank money has more than trebled over the last third of the 20th century
Future Innovations and the Income Velocity of Central Bank Money
Looking ahead to the 21st century, we can expect continued increases in velocity. Substitutes for, and economizers of, current money assets, and increasingly sophisticated money transfer systems, all are on the horizon.
But will this just be a case of "déjà vu, all over again," another episode of innovation increasing the income velocity of central bank money, making mischief with quantitative monetary targets? Perhaps not, for another possibility must be recognized. The kernel of the money question beginning to emerge on the 21st century horizon is not just about the predictability of changes in velocity, or even the instability induced by ever higher velocity. Rather, what may be new and different about the 21st century is the possibility that the velocity of central bank money might approach infinity--that is, that there will be no appreciable domestic demand at all for central bank money--whether currency or banks' deposit balances at reserve banks.
Discussions of smart card and Internet moneys hint at this radically new monetary future. One focus is on the degree to which value embedded in smart-card memories will be the liability of commercial firms or of financial institutions, and whether traditional regulations such as reserve requirements and capital ratios might extend to smart cards, whoever issues them. These questions, while important in the short-run, may be largely beside the point in the long run. Reserve requirements already are becoming a dead issue, killed by technology and competition. Capital ratio requirements are meeting the same fate, from the same forces: To the extent capital ratios might be made more onerous than the value of the safety net services they buy, they are unlikely to survive in the long run.
Smart card and Internet moneys must meet quality control standards in some form, of course, either from safety net supervision or from pressures of customers and competitors in the market. Safety and soundness always will be relevant to customers' choices among moneys. Similarly, the relative quantities of these moneys will be controlled by competing with alternative monetary assets and transfer mechanisms like credit cards and debit cards, as well as paper checks and electronic transfers of account balances. Thus, regulation of new electronic moneys is a doubtful source of future demand for central bank money.
Issuance of successful electronic moneys by the central bank itself would ensure a continuing demand for central bank liabilities. This possibility cannot be ruled out for the distant future. For now, however, neither government regulation of private issuers nor direct government issuance of money seems likely to create significant demand for central bank money. In the retail payments industry, the question is whether customers will continue to demand central bank currencies, or will replace such currencies with electronic moneys. The answer will depend on customers' evaluations of relative features and costs, some of which can be enumerated now. The new technologies all have the apparent advantages of no bulk, perfect divisibility, and delivery versus payment. Float is positive on credit cards, zero on debit cards, but, in a sense, negative on smart card and Internet payments balances.
What does currency offer in place of these features? Currency is a relatively bulky medium of exchange, maintained by networks of ATMs, branch offices, and armored trucks. It is issued only in fixed denominations, and bears no nominal rate of return. It's advantages come from allowing delivery versus final payment, providing immediate finality, as well as allowing anonymity for some users. On balance, it seems quite likely that electronic moneys could supplant currency in a considerable portion of legal, retail transactions.
Complete substitution of electronic moneys for currency in domestic use would leave a substantial quantity of central bank money outstanding. Foreign holdings, remember, are estimated to represent about two-thirds of the outstanding value of U.S. currency. The durability of this demand may depend more on the relative qualities of U.S. and foreign monetary management in the next century than on the relative costs and features of currency and its electronic substitutes. It is not at all clear, however, even if the United States were assured of another century of foreign demand for central bank money, that controlling the supply of currency to foreign holders is an efficacious method of conducting domestic monetary policy.
Another hint of a radically new 21st century monetary future comes from looking at the demand for central bank money by depository institutions. If technology and competition eliminate demands for currency by the general public, then depository institutions' derived demand for vault cash will wither. Moreover, technology, competition, and regulatory actions, already have removed a substantial part of the demand for reserve bank balances to satisfy reserve requirements. It seems possible that this process could continue until virtually no bank in the United States was constrained by reserve requirement regulations as currently structured. Therefore, low reserve requirements may be just as untenable tomorrow as high requirements proved to be in the past.
Reserve requirements are not the only reason for maintaining an account balance with a reserve bank. Many banks maintain clearing balances that serve two business purposes. They provide a cushion that protects against daylight and overnight overdrafts, and they receive a market-based rate of return, although it can be used only to pay for financial services provided by the reserve banks.
Neither reason for holding a clearing balance is a very robust source of demand for central bank money. Avoidance of overdrafts can be achieved in other ways. One is to apply information technology to the sequencing of debits and credits during a day to minimize daylight overdrafts and avoid surprise debits at the end of a day. Another is to organize and participate in multilateral clearing and net settlement arrangements for money and securities transfers that could substitute for reserve bank services, reducing the need for a cushion.
These and other alternatives to holding balances may not be especially attractive today because banks tend to use the reserve banks' priced payment services in sufficient volume for earnings credits on clearing balances to be valuable. Over time, however, definitive paper instruments will lose market share, eventually rendering check and noncash collection services obsolete. Moreover, commercial competitors are likely to continue making inroads in the growing market for automated clearing house services that once was almost the exclusive domain of the reserve banks.
It is not really a complete flight of fancy to foresee a possible outcome in which central bank money becomes insignificant in the domestic economy. The public may find commercially provided electronic money attractive as a replacement for currency. Reserve requirements are not likely to provide a solid floor under the demand for reserve bank money by depository institutions. Finally, demand for clearing balances at the reserve banks could decline as earnings credits become less valuable.
What Role Remains for the Central Bank?
Even with little demand by the public to hold central bank liabilities, central banks remain the only source of the national currency units that are required to settle domestic tax obligations. The Federal Reserve, and every other central bank of which we are aware, provides settlement finality as a payments service. Final settlement represents an ultimate, official guarantee of values exchanged by reserve bank depositors, thereby allowing banks to verify exchanges of value among their customers. Finality may be rendered on a gross basis, as the reserve banks do in making immediate, irrevocable Fedwire transfers, or on a net basis, as the reserve banks do in settling the zero-sum end-of-day positions of its' depositors who belong to a multilateral clearing house arrangement like CHIPS.
Central banks effect settlement when they post irrevocable debits and offsetting credits to two or more depository institutions' account balances. In the United States, those account balances must be zero or positive at the end of each day, and typically are in the $30 billion range in the aggregate. At the Bank of England, in contrast, aggregate balances are close to zero at the end of each day. This highlights the fact that overnight balances are not necessary, either in the aggregate or for an individual depository institution, as long as the supply of intraday credit is sufficient to accommodate mismatched flows of depositors' receipts and payments without payments gridlock. Intraday credit might come from the central bank, as in real-time gross settlement systems like Fedwire, of from other network participants, as in batched settlement systems like CHIPS.
Looked at in this way, the settlement function of the central bank will continue even though central bank money is not widely held. Saying the same thing another way, central bank money may not be used as an asset in which to store value overnight and longer, but it still will be required as a vehicle for transferring value during a day.
The monetary policy function of the central bank--maintaining constant purchasing power of the standard of value, or unit of account--still must be fulfilled even if central bank liabilities are not a common form of balances denominated in the national currency units. Even if the long-run equilibrium value of the domestic demand for monetary base approached zero, the policy authority could still control the terms on which payments system institutions would acquire funds to pay-off debts to the central bank contracted during the day, and the terms on which institutions could rid themselves of excess funds accumulated during the day to maintain that zero balance. Treasury and central bank payments and receipts would seem to be the critical factors in determining whether depositories developed an aggregate net debt or net credit position during a day, just as is the case today. The mechanisms a central bank might find useful in maintaining zero liabilities to the private sector, of course, might be quite different from those now used to maintain more than $400 billion in liabilities, but the principle involved--zero excess supply and demand at a desired level of a money market interest rate--would seem to be essentially the same.
We have just outlined how the settlement and monetary policy roles of the central bank might be carried on, even in the absence of a demand for central bank money in the next century. “Might be carried on” is different, of course, from “will be carried on,” and we confess to uncertainty about the impacts of less-readily analyzed pressures for change that might accompany declining demand for central bank money.
For example, a central bank with few liabilities also has few assets, and therefore little interest income. In the United States, the interest income of the reserve banks is one of the protections of central bank independence. It is their interest income that supports the operations of the Federal Open Market Committee, including staff monetary policy research and analysis at each of the reserve banks and at the Board of Governors.
Also, a central bank with minimal assets and liabilities has limited ability to act as a buffer between the Treasury and the financial system. Payment-day for each fresh Treasury issue would see depositories thrown into daylight overdraft positions until they could (a) borrow an equivalent amount longer term from the central bank, (b) accept Treasury deposits of equivalent amount, or (c) receive the proceeds of reserve bank purchases of an equivalent value of Treasury (or other) securities.
Conclusion
While we can foresee the possibility that central bank money will not be widely held in the 21st century, we cannot predict either that it will happen, or what the consequences of its happening would be for methods of settling payments and mechanisms for managing the purchasing power of the unit of account. Our uncertainty is relieved, however, by seeing one direction in which academic research has been moving. The general topic of “free banking” can be thought of as dealing with how an economic and financial system would operate in the absence of state interventions such as a central bank. Alternative definitions of “free banking” are being used, it's true, ranging from a money industry in which banks operate without reference to a common unit of account, to a money industry not much different from the long-run situation we have been assuming, in which financial markets avoid all regulations that provide no quid pro quo. Nonetheless, research evolving in these directions is precisely what is needed for the next century, when there is a good chance that central bank money will not be in much demand.
[1] Negotiable Orders of Withdrawal.
[2] This can be little more than a suspicion, however, given the apparent inconsistencies displayed by alternative methods of estimating the time path of currency held abroad. (See, e.g., Porter & Judson, and Feige, WEA Meetings, July 1995.)
Prepared for the Cato Institute's 14th Annual Monetary Conference, May 23, 1996, Washington, D.C.