By JKH (cross posted at Monetary Realism)
Loans create deposits. We’ve heard it many times now. But how well is it understood? The phrase is typically invoked accurately, in conjunction with a rejection of the ‘money multiplier’ fable found in economic textbooks. From an operational perspective, banks do not “lend reserves” to their non-bank customers. “Loans create deposits’ is an operation in endogenous money. And where central banks impose a level of required reserves based on deposits, the timing of the demand for and supply of reserves in respect of such a requirement follows the creation of the deposit – it does not precede it. The money multiplier story is bunk. And ‘loans create deposits’ is correct as an observation.
Nevertheless, there is a larger context for deposits, which includes their fate after they have been created. Deposits are used to repay loans, resulting in the ‘death’ of both loan and deposit. But there is more. As part of the birth/death analogy, there is the lifetime of loans and deposits to consider. This sequence of birth, life, and death in total may be helpful in putting ‘loans create deposits’ into a broader context. There is potential for confusion if ‘loans create deposits’ is embraced too enthusiastically as the defining characteristic, without considering the full life cycle of loans and deposits. Indeed, we shall see further below that ‘deposits fund loans’ is as true as ‘loans create deposits’ and that there is no contradiction between these two things.
The monetary system and the financial system are constructions of double entry accounting. It has been this way for a long time. This did not start in 1971. The fact that there was gold serving as a fixed value backstop for certain monetary assets shouldn’t obscure the fact that a monetary system is a fiat construction at its foundation. Gold at one time was a hard constraint on the behavior of the monetary authorities. But the authorities will inevitably create paradigms of operational constraint and guidelines in any monetary system. These restrictions include central bank balance sheet constraints (e.g. gold backing; Treasury overdraft constraints; supply and pricing of bank reserves that are consistent with the monetary policy interest rate target) and other guidelines (such as the reaction function of the policy rate to various measures of inflation, output, or employment). The full category of potential constraints is broad and varied. But none of this alters the fact that a monetary system is basically a bookkeeping device for the intermediation of real economic activity. It is a construct that enables moving beyond a barter economic system that can only be imagined as a counterfactual.
The Choice for Banking
Starting from this monetary bookkeeping foundation, a fundamental choice exists. Will the system include a competitive banking sector? More broadly, will financial capitalism exist in substance and form? Will there be competition? Within this landscape, will there be more than one bank? While a banking singularity (a single, concentrated, nationalized institution) is usually considered to be non-pragmatic, it serves as a useful theoretical reference point for understanding how banks actually work. The competitive framework that is often taken for granted is in fact a choice for banking system design – including the presence of a reserve system that enables active management of individual bank balance sheets.
‘Loans Create Deposits’
When we say ‘loans create deposits’, we mean at least that the marginal impact of new lending will be to create a new asset and a new liability for the banking system – typically for the originating lending bank at first. A bank makes a loan to a borrowing customer. That is a debit under bank assets. Simultaneous, it credits the deposit account of the same customer. That is a new bank liability. Both of those accounting entries represent increases in their respective categories. This is operationally separate from any notion of reserves that may be required in association with the creation of bank deposits.
In another version of the same lending transaction, the lending bank presents the borrower with a cheque or bank draft. The lending bank debits the borrower’s loan account and credits a payment liability account. The bank’s balance sheet has grown. The borrower may then deposit that cheque with a second bank. At that moment, the balance sheet of the second bank – the deposit issuing bank – grows by the same amount, with a payment due asset and a deposit liability. This temporary duplication of balance sheet growth across two different banks is captured within the accounting classification of bank ‘float’. The duplication gets resolved and eliminated when the deposit issuing bank clears the cheque back to the lending bank and receives a reserve balance credit in exchange, at which point the lending bank sheds both reserve balances and its payment liability. The end result is that the system balance sheet has grown by the amount of the original loan and deposit. The loan has created the deposit, although loan and deposit are domiciled in different banks. The system has expanded in size. The growth is now reflected in the size of the deposit issuing bank’s balance sheet, with an increase in deposits and reserve balances. The lending bank’s balance sheet size is unchanged from the start (at least temporarily), with loan growth offset by a reserve balance decline.
In this latter example, it is possible and even likely, other things equal, that the lending bank additionally will seek to borrow new funding from wholesale money markets and that the deposit issuing bank will lend funds into this market. This is a natural response to the respective change in reserve distribution that has been created momentarily for the two banks. Without further action, the lending bank has lost reserves and the deposit bank has gained reserves. They may both seek to normalize these respective reserve positions, other things equal. Adjusting positions through money market operations is a basic function of commercial bank reserve management. Thus, this example features the core role of bank reserves in clearing a payment from one bank to another. The final resolution of positions in this case is that the balance sheets of both banks will have expanded, indirectly connected through money market transactions that follow on from the initial ‘loans create deposits’ transaction. However, this too may be a temporary situation, as the original transaction involving two different banks will inevitably be followed up by further transactions that shift bank reserves between various bank counterparties and in various directions across the system.
The Money Multiplier Fable
The money multiplier story – a fable really – claims that banks expand loans and deposits on the basis of a central bank function that gradually feeds reserves to banks, allowing them to expand their balance sheets with new loans and reservable deposits – according to reserve ratios that bind the pace of that expansion according to the reserves supplied. This is entirely wrong, of course. In fact, bank balance sheet expansion occurs largely through the endogenous process whereby loans create deposits. And central banks that impose reserve requirements provide the required reserve levels as a matter of automatic operational response – after the loan and deposit expansion that generates the requirement has occurred. The multiplier fable describes a central bank with direct exogenous control over bank expansion, based on a reserve supply function – which is a fiction. The facts of endogenous money creation have been demonstrated by empirical studies going back decades. Moreover, the facts are obvious to anybody who has actually been involved with or closely studied the actual reserve management operations of either a commercial bank or a central bank. In truth, no empirical ‘study’ is required – the banking world operates this way on a daily basis – and it is absurd that so many economics textbooks make up stories to the contrary. The truth of the ‘loans creates deposits’ meme is pretty well understood now – at least by those who take the time to learn the facts about it.
Central Bank Reserve Injections
A central bank that imposes a reserve requirement will follow up new deposit creation with a system reserve injection sufficient to accommodate the requirement of the individual bank that has issued the deposit. The new requirement becomes a targeted asset for the bank. It will fund this asset in the normal course of its asset-liability management process, just as it would any other asset. At the margin, the bank actually has to compete for funding that will draw new reserve balances into its position with the central bank. This action of course is commingled with numerous other such transactions that occur in the normal course of reserve management. The sequence includes a time lag between the creation of the deposit and the activation of the corresponding reserve requirement against that deposit. Thus, there is a lag between two system growth impulses – ‘loans create deposits’ as the endogenous feature and a subsequent central bank reserve injection as an exogenous follow up. The required reserve injection is typically small by comparison, according to the reserve ratio. The central bank can provide the reserves in different ways, such as by purchasing bonds or by conducting system repurchase operations with investment dealers. In the case of either bond purchases or system repurchase agreements, additional system deposits might be created when the end seller (or lender) of the bonds is a non-bank. And that second order creation of deposits may be reservable as well. But what might appear to be a potentially infinite series of reserve injections is in fact highly controlled in the real world – because the reserve ratio is relatively small. Some countries such as Canada have no such required reserve ratio. Indeed, the case of zero required reserves nicely emphasizes the nature of the money multiplier as an annoying analytical error and distraction from accurate comprehension of how banks actually work. But as a separate point, central bank injections of required reserves illustrate how not all deposits are necessarily created by commercial bank loans. ‘Loans create deposits’ is true, but not exclusive. This aspect is made clear also by the example of central bank ‘quantitative easing’, noted further below.
The Growth Dynamic
The ‘loans create deposits’ meme is best understood as a balance sheet growth dynamic, distinct from any reserve effect that might occur as part of an associated interbank clearing transaction at the time (e.g. the second example above) or as part of a deposit ratio requirement that might be activated at a later date. The banking system can be visualized in continuous time, punctuated by discrete banking transactions that are reflected as accounting entries. If one divides time into very small time intervals, individual banking transactions can be isolated as the only transactions that occur during a given interval of time. Thus, the growth dynamic of ‘loans create deposits’ can be conceived of as an instantaneous balance sheet expansion at the point of corresponding accounting entries. As noted in the examples above, this expansion may then migrate across individual banks when the lending and deposit issuing bank are different.
‘Deposits Fund Loans’
Some interpretations of the ‘loans create deposits’ meme overreach in their desired meaning. The contention arises occasionally that ‘loans create deposits’ means banks don’t need deposits to fund loans. This is entirely false. This is the point that requires emphasis in this essay.
There is no inconsistency between the idea that ‘loans create deposits’ and the idea that banks need deposits to fund loans. Bank balance sheet management must respond to both growth dynamics and steady state conditions in the dimension of nominal balance sheet size. A bank in theory can temporarily be at rest in terms of balance sheet growth, and still be experiencing continuous shifting in the mix of asset and liability types – including shifting of deposits. Part of this deposit shifting is inherent in a private sector banking system that fosters competition for deposit funding. The birth of a demand deposit in particular is separate from retaining it through competition. Moreover, the fork in the road that was taken in order to construct a private sector banking system implies that the central bank is not a mere slush fund that provides unlimited funding to the banking system. In fact, active liability management is important in private sector banking – in the system we actually have. Other systems have been proposed, in which central banks intervene in some way to adjust the landscape of competitive liability management (e.g. the Chicago Plan; full reserves) or to subsume this competition more comprehensively (e.g. the MMT Mosler plan). These are ideas for significant change that should not be confused with the characteristic of competitive banking as it now exists. Some analysts tend toward language that conflates factual and counterfactual cases in this regard. To repeat – bank liability management is very competitive in the system we have, by design. The ‘loans create deposits’ meme, while true, only touches on this competitive dynamic.
We note again that loans are not the sole source of deposit creation. A commercial bank’s purchase of securities from a non-bank will typically result in new deposit creation somewhere in the system. There are cases where deposit creation results from other liability or equity conversion – commercial bank debt redemption and stock buybacks are examples of this. Existing fixed term deposits can convert to demand deposits and vice versa. And central bank quantitative easing most often results in new deposit creation – because the bonds that the central bank purchases are typically sourced from non-bank portfolios, and exchanged for deposits. Nevertheless, ‘loans creates deposits’ is a reasonable reference point and standard for the process of deposit creation.
Bank Asset-Liability Management
The ‘loans create deposits’ dynamic comprises the production of much of the money that serves as a basic source of liquidity in a monetary economy. The originating accounting entries are simple – a loan asset and a deposit liability. But this is only the start of the story. Commercial bank ‘asset-liability management’ functions oversee the comprehensive flow of funds in and out of individual banks. They control exposure to the basic banking risks of liquidity and interest rate sensitivity. Somewhat separately, but still connected within an overarching risk management framework, banks manage credit risk by linking line lending functions directly to the process of internal risk assessment and capital allocation. Banks require capital – especially equity capital – to take risk – and to take credit risk in particular.
Interest rate risk and interest margin management are critical aspects of bank asset-liability management. The ALM function provides pricing guidance for deposit products and related funding costs for lending operations. This function helps coordinate the operations of the left and the right hand sides of the balance sheet. For example, a central bank interest rate change becomes a cost of funds signal that transmits to commercial bank balance sheets as a marginal pricing influence. The asset-liability management function is the commercial bank coordination function for this transmission process, as the pricing signal ripples out to various balance sheet categories. Loan and deposit pricing is directly affected because the cost of funds that anchors all pricing in finance (e.g. the fed funds rate) has been changed. In other cases, a change in the term structure of market interest rates requires similar coordination of commercial bank pricing implications. And this reset in pricing has implications for commercial bank approaches to strategies and targets for the compositional mix of assets and liabilities.
The life of deposits is more dynamic than their birth or death. Deposits move around the banking system as banks compete to retain or attract them. Deposits also change form. Demand deposits can convert to term deposits, as banks seek a supply of longer duration funding for asset-liability matching purposes. And they can convert to new debt or equity securities issued by a particular bank, as buyers of these instruments draw down their deposits to pay for them. All of these changes happen across different banks, which can lead to temporary imbalances in the nominal matching of assets and liabilities, which in turn requires active management of the reserve account level, with appropriate liquidity management responses through money market operations in the short term, or longer term strategic adjustment in approaches to loan and deposit market share. The key idea here is that banks compete for deposits that currently exist in the system, including deposits that can be withdrawn on demand, or at maturity in the case of term deposits. And this competition extends more comprehensively to other liability forms such as debt, as well as to the asset side of the balance sheet through market share strategies for various lending categories. All of this balance sheet flux occurs across different banks, and requires that individual banks actively manage their balance sheets to ensure that assets are appropriately and efficiently funded with liabilities and equity.
In examining all of these effects, it is helpful to consider the position of the banking system in its totality, in conjunction with the position of individual banks that constitute the whole. For example, the US commercial banking system is composed of thousands of individual banks. Between discrete ‘loans create deposits’ events, the banking system is in continuous balance sheet churn. Specifically, deposits are moving back and forth between individual banks, as a matter of normal payment system operations. They are also moving and inter-converting in the form of term deposits at both the retail and wholesale level. This overall liquidity churn feeds economic activity of all sorts, where households, businesses, and governments are making payments to each other for various goods and services and other types of transactions, and are making choices about the portfolio structure of their liquid assets. This is the core liquidity provided by the banks to their customers. And this is the stuff that involves a good deal of transferring of reserves back and forth between banks, in order to affect accounting completion of balance sheets that are in continuous flux in size and composition.
Bank Reserve Management
The ultimate purpose of reserve management is not reserve positioning per se. The end goal is balance sheets that are in balance, institution by institution – and where deposits fund loans, alongside various other asset-liability matching configurations. The reserve system records the effect of this balance sheet activity. The reserve account is the inverse exogenous money image of the nominal configuration of the rest of the balance sheet. The balance sheet requires asset liability management coordination in order to match up assets and liabilities both in nominal terms and in a way that is financially effective. And even if loan books remain temporarily unchanged, all manner of other banking system assets and liabilities may be in motion. This includes securities portfolios, deposits, debt liabilities, and the status of the common equity and retained earnings account. And of course, loan books don’t remain unchanged for very long, in which case the loan/deposit growth dynamic comes directly into play on a recurring basis.
In summary, the original connection by which deposits are created by loans typically disappears at some point following deposit creation – at the micro bank level and/or the macro system level. The original demand deposits associated with specific loan creation become commingled as they move back and forth between different banks. And they not only move between banks, but they can change in form within any bank. They can be converted into term deposits or other funding forms such as bank debt or common and preferred stock. The task of dealing with this compositional flux falls under the joint coordination of bank asset-liability management and reserve management. The overarching point of observation is that both system growth and system competition for existing balance sheet composition are in constant operation. ‘Loans create deposits’ only describes the marginal growth dynamic at the inception of deposit creation. ‘Deposits fund loans’ is the more apt description that applies to a good portion of what constitutes ongoing balance sheet management in competitive banking.