Money is usually treated as a monetary phenomenon and accounting as a system of financial reporting.
The two are often analyzed separately.
Yet modern money is created, held, transferred, and extinguished through balance sheets.
Commercial banks create deposits when they extend credit. Central banks issue reserve balances and currency. Firms and households hold claims on banks that function as means of payment. Financial institutions continuously classify assets according to liquidity, maturity, risk, and accounting treatment.
For this reason, the monetary system is also an accounting system.
A simplified bank balance sheet can be written as
\[ A=L+E, \]
where \(A\) denotes assets, \(L\) liabilities, and \(E\) equity.
Among the liabilities of a commercial bank are deposits.
For the non-bank sector, those deposits are money.
This produces an immediate connection between accounting and monetary theory:
\[ \boxed{ \text{Bank accounting} \rightarrow \text{balance-sheet capacity} \rightarrow \text{credit creation} \rightarrow \text{deposit creation} \rightarrow \text{money supply} } \]
Accounting rules do not mechanically determine the quantity of money.
But they can influence the balance-sheet constraints and classifications through which money creation takes place.
Consider a commercial bank making a loan of €100,000.
At the moment the loan is granted, the bank records an asset:
\[ \text{Loan}=+100{,}000 \]
and simultaneously records a deposit liability:
\[ \text{Deposit}=+100{,}000. \]
The transaction can be represented as
\[ \Delta A=+100{,}000, \]
\[ \Delta L=+100{,}000. \]
The bank's balance sheet expands.
From the customer's perspective, the new deposit can be used to make payments.
Thus, under a deposit-based monetary system,
\[ \boxed{ \text{Bank credit creation} \rightarrow \text{deposit creation} } \]
The monetary system therefore cannot be completely separated from accounting representation.
Money exists on balance sheets.
It is tempting to think that accounting simply records a loan after the bank has already made an economic decision.
But accounting rules affect how the consequences of that decision appear on the bank's balance sheet.
Suppose the economic position of a bank is represented by
\[ E, \]
while the accounting system reports
\[ A=\mathcal{A}(E), \]
where \(\mathcal{A}\) is the accounting operator.
This is the same general structure developed in the Accounting Bias framework.
The reported balance sheet is not economic reality directly.
It is a representation of economic reality produced through rules governing
\[ \text{recognition}, \quad \text{measurement}, \quad \text{classification}, \quad \text{timing}, \quad \text{estimation}. \]
In banking, those rules may affect quantities that matter directly for financial intermediation.
Examples include
\[ \text{loan values}, \]
\[ \text{securities values}, \]
\[ \text{impairment allowances}, \]
\[ \text{profit}, \]
\[ \text{equity}, \]
\[ \text{liquidity classifications}. \]
These accounting quantities can influence how much additional balance-sheet expansion a bank is willing or able to undertake.
Recognition is the first channel.
Suppose an accounting framework permits a particular financial claim to be recognized as an asset.
Then the balance sheet contains
\[ A_i>0. \]
If the same economic claim were not recognized, the reported amount would instead be
\[ A_i=0. \]
This does not automatically create money.
But recognition may alter the institution's reported resources and financial position.
The important distinction is therefore:
\[ \text{asset recognition} \neq \text{money creation}. \]
Rather,
\[ \text{asset recognition} \rightarrow \text{reported balance-sheet structure} \rightarrow \text{possible effects on financial capacity}. \]
The monetary importance depends on what the recognized asset can be used for.
If an asset is regarded as liquid, saleable, pledgeable, or acceptable as collateral, its recognition may be particularly important.
The relationship becomes especially clear when accounting practice causes a larger set of assets to be recognized or classified as highly liquid.
Let
\[ LA \]
denote reported liquid assets.
Then a more permissive recognition or classification rule may produce
\[ LA^{(1)}>LA^{(0)}. \]
This alone does not imply
\[ M^{(1)}>M^{(0)}, \]
where \(M\) denotes the money supply.
But it may affect the institution's willingness or capacity to expand liabilities.
Suppose a bank internally targets a liquidity ratio
\[ \lambda = \frac{LA}{D}, \]
where \(D\) denotes deposits.
If management requires
\[ \lambda\geq\bar{\lambda}, \]
then the maximum level of deposits consistent with the target is
\[ D_{\max} = \frac{LA}{\bar{\lambda}}. \]
If reported liquid assets rise,
\[ LA\uparrow, \]
then mechanically,
\[ D_{\max}\uparrow. \]
In this simplified representation, accounting treatment can therefore influence the balance-sheet capacity associated with deposit creation.
The causal chain becomes
\[ \boxed{ \text{Accounting treatment} \rightarrow \text{reported liquid assets} \rightarrow \text{liquidity constraint} \rightarrow \text{credit capacity} \rightarrow \text{deposit creation} } \]
This is a more defensible statement than saying that recognizing liquid assets automatically increases money.
Measurement provides a second channel.
Suppose a bank holds a security whose accounting value is
\[ A_t. \]
Under one accounting treatment it may remain close to historical or amortized cost.
Under another, it may be measured at a current market-based amount:
\[ A_t=P_t. \]
If market prices rise,
\[ P_t\uparrow, \]
then reported asset values may increase.
Depending on where the revaluation is recognized, the increase may affect
\[ \text{profit}, \]
\[ \text{equity}, \]
or
\[ \text{other comprehensive income}. \]
In abstract form,
\[ \Delta P \rightarrow \Delta A \rightarrow \Delta E. \]
If equity is an economically relevant constraint on balance-sheet expansion, then accounting measurement may indirectly affect credit creation.
For example, suppose a bank follows an internal leverage constraint
\[ \frac{A}{E} \leq \bar{\ell}. \]
Then
\[ A_{\max} = \bar{\ell}E. \]
If an accounting remeasurement increases reported equity,
\[ E\uparrow, \]
then
\[ A_{\max}\uparrow. \]
The institution may consequently have greater capacity to expand lending.
Again, the relationship is indirect:
\[ \text{Accounting measurement} \rightarrow \text{reported equity} \rightarrow \text{balance-sheet capacity} \rightarrow \text{potential credit creation}. \]
Accounting may also restrict monetary expansion.
Consider loan impairment.
A bank initially recognizes a loan at
\[ L. \]
Suppose expected credit losses are recognized as an allowance
\[ EL. \]
The net carrying amount becomes
\[ L_{\text{net}} = L-EL. \]
The associated impairment expense reduces profit and therefore, other things equal, equity.
Thus,
\[ EL\uparrow \]
may imply
\[ \text{Profit}\downarrow \]
and subsequently
\[ E\downarrow. \]
If the bank's lending activity is constrained by capital or internal risk limits, then
\[ E\downarrow \rightarrow \text{credit capacity}\downarrow. \]
Because bank lending can create deposits,
\[ \text{credit growth}\downarrow \rightarrow \text{deposit-money growth}\downarrow. \]
Accounting conservatism can therefore potentially operate as a contractionary balance-sheet mechanism.
Conversely, delayed recognition of deterioration may temporarily leave reported equity higher and permit more aggressive balance-sheet expansion.
This introduces the timing dimension.
Two accounting systems may ultimately recognize the same economic loss, but at different dates.
Suppose the economic deterioration occurs at time \(t\).
One accounting system recognizes the loss immediately:
\[ Loss_t. \]
Another recognizes it later:
\[ Loss_{t+k}. \]
Then during the interval
\[ [t,t+k), \]
the second system reports higher assets, profit, or equity than the first.
This creates a temporal accounting difference:
\[ AB_t^{T} = A_t-E_t. \]
In banking, such timing differences may matter because lending decisions are made using current balance sheets.
A delayed loss can therefore affect real financial decisions before eventual recognition occurs.
The sequence can be written as
\[ \text{delayed recognition} \rightarrow \text{temporarily higher equity} \rightarrow \text{temporarily larger lending capacity} \rightarrow \text{higher deposit creation}. \]
The reverse is possible when losses are recognized earlier.
Accounting timing can therefore influence the path of monetary expansion even when cumulative long-run recognition eventually converges.
This mechanism becomes especially important over the business cycle.
Suppose asset prices rise during an economic expansion.
If accounting values respond strongly to market prices, then
\[ P_t\uparrow \rightarrow A_t\uparrow. \]
Higher asset values may improve reported financial ratios and strengthen perceived collateral values.
This may support additional lending:
\[ Credit_t\uparrow. \]
Additional loans create additional deposits:
\[ Deposits_t\uparrow. \]
The process may therefore reinforce the expansion:
\[ \boxed{ P\uparrow \rightarrow A\uparrow \rightarrow \text{financial capacity}\uparrow \rightarrow Credit\uparrow \rightarrow Deposits\uparrow } \]
During a downturn, the mechanism may reverse:
\[ P\downarrow \rightarrow A\downarrow \rightarrow E\downarrow \rightarrow Credit\downarrow. \]
Accounting measurement can therefore contribute to financial procyclicality when reported values move with market conditions and those values affect lending decisions.
A related channel operates through collateral.
Suppose a borrower owns an asset with recognized value
\[ C. \]
A lender permits borrowing up to a fraction
\[ \theta \]
of collateral value.
Then
\[ Loan_{\max} = \theta C. \]
If accounting or valuation practices increase recognized collateral value,
\[ C\uparrow, \]
then
\[ Loan_{\max}\uparrow. \]
If that loan is created by a commercial bank, new deposits may be created simultaneously.
Thus,
\[ \boxed{ \text{Asset valuation} \rightarrow \text{collateral value} \rightarrow \text{borrowing capacity} \rightarrow \text{bank credit} \rightarrow \text{money creation} } \]
This channel need not operate through the bank's own accounting alone.
Accounting practices applied to borrowers can also influence lending.
There is a deeper version of the argument.
Not all assets are equally monetary.
Consider a continuum from very illiquid productive assets to highly liquid financial claims.
One could represent monetary proximity as
\[ \mu_i\in[0,1], \]
where
\[ \mu_i=0 \]
represents an economically useful but highly illiquid asset, while
\[ \mu_i=1 \]
represents something functioning almost like money.
An institution's liquidity-weighted asset stock could then be written as
\[ LA^{*} = \sum_i \mu_i A_i. \]
Accounting classification affects the observable \(A_i\), while institutional rules affect the effective monetary weight \(\mu_i\).
If accounting practice recognizes more assets with high liquidity characteristics,
\[ LA^{*}\uparrow. \]
This may increase the ease with which balance-sheet assets can be transformed into payment capacity.
One can therefore distinguish between
\[ \text{money}, \]
\[ \text{money-like claims}, \]
and
\[ \text{illiquid assets}. \]
Accounting practices help define the reported boundaries among these categories.
This is an important limitation.
Suppose a company reclassifies a factory as a current asset.
That accounting entry does not suddenly make the factory a medium of exchange.
Economic liquidity has institutional and market dimensions.
True liquidity depends on characteristics such as
\[ \text{market depth}, \quad \text{convertibility}, \quad \text{settlement speed}, \quad \text{price stability}, \quad \text{counterparty acceptance}. \]
Therefore,
\[ \text{accounting liquidity} \neq \text{economic liquidity}. \]
The Accounting Bias framework is useful precisely because it allows these two concepts to be separated.
Let
\[ A_L \]
represent accounting classification of liquidity, and
\[ E_L \]
the underlying economic liquidity.
Then
\[ AB_L = A_L-E_L. \]
An aggressive liquidity classification could therefore create a positive liquidity-related Accounting Bias:
\[ A_L>E_L. \]
In such a case, the financial system might behave as if it contained more readily mobilizable assets than actually exists under stress.
A further distinction is necessary.
Accounting can affect both
\[ \text{actual monetary capacity} \]
and
\[ \text{statistically measured money}. \]
These are not always the same problem.
Official monetary aggregates classify financial claims according to institutional and liquidity characteristics.
If classification conventions change, a financial instrument may move into or out of a monetary aggregate.
Then measured money may change even if the underlying economic arrangements have barely changed.
Conceptually,
\[ M = \sum_i \omega_i F_i, \]
where \(F_i\) denotes financial claims and \(\omega_i\) indicates whether or to what degree each claim is counted as money.
A classification change can alter
\[ \omega_i. \]
Therefore,
\[ \Delta M \]
can sometimes reflect a change in statistical or accounting boundaries rather than new purchasing power.
This is another form of measurement bias.
The framework can therefore be extended.
Let economic monetary capacity be
\[ M^{E}, \]
and measured monetary capacity be
\[ M^{A}. \]
One can define a monetary representation gap:
\[ MB = M^{A}-M^{E}. \]
This should not necessarily be interpreted as an error.
Rather, it captures the divergence between the institutional measure of money and some underlying economic notion of liquidity or payment capacity.
The same logic used throughout Accounting Bias appears again:
\[ \text{economic financial structure} \rightarrow \text{classification and measurement} \rightarrow \text{reported monetary structure}. \]
Thus,
\[ \boxed{ M^{A} = \mathcal{M} \left( M^{E}, R, V, T, C \right) } \]
where \(R\) represents recognition rules, \(V\) valuation rules, \(T\) timing, and \(C\) classification conventions.
The relationship can be summarized in a stylized banking model.
Let
\[ L_t \]
denote bank loans,
\[ D_t \]
deposits,
\[ LA_t \]
liquid assets, and
\[ K_t \]
reported equity.
Suppose lending capacity depends on both liquidity and capital:
\[ L_t^{\max} = f(LA_t,K_t). \]
Assume
\[ \frac{\partial L^{\max}}{\partial LA}>0 \]
and
\[ \frac{\partial L^{\max}}{\partial K}>0. \]
Accounting rules affect both variables:
\[ LA_t = \mathcal{A}_L(E_t), \]
\[ K_t = \mathcal{A}_K(E_t). \]
Then
\[ L_t^{\max} = f \left[ \mathcal{A}_L(E_t), \mathcal{A}_K(E_t) \right]. \]
If deposit creation approximately follows new bank lending,
\[ \Delta D_t \approx \Delta L_t, \]
then accounting influences the monetary system through
\[ \boxed{ \mathcal{A} \rightarrow L^{\max} \rightarrow L \rightarrow D \rightarrow M } \]
This is the core theoretical proposition.
We can define an Accounting Expansion Channel.
Suppose a new accounting treatment causes
\[ \Delta LA>0 \]
or
\[ \Delta K>0. \]
If financial constraints are binding, then
\[ \Delta L^{\max}>0. \]
If banks use the additional capacity,
\[ \Delta L>0. \]
Since bank lending creates deposits,
\[ \Delta M>0. \]
Thus:
\[ \boxed{ \Delta \mathcal{A} \rightarrow \Delta LA,\Delta K \rightarrow \Delta L \rightarrow \Delta M } \]
The accounting rule has not created money directly.
It has altered a constraint governing institutions that can create money.
That distinction is essential.
The opposite is equally possible.
Suppose accounting rules accelerate impairment recognition.
Then
\[ Loss\uparrow, \]
\[ K\downarrow. \]
If bank balance sheets are constrained,
\[ L^{\max}\downarrow. \]
Lending may contract:
\[ \Delta L<0. \]
Loan repayment or deleveraging can reduce deposits:
\[ \Delta D<0. \]
Thus,
\[ \boxed{ \text{Accounting loss recognition} \rightarrow \text{equity contraction} \rightarrow \text{credit contraction} \rightarrow \text{monetary contraction} } \]
Accounting therefore has potentially asymmetric monetary effects across expansions and recessions.
This suggests a broader interpretation.
Monetary theory often focuses on variables such as
\[ r, \]
the policy interest rate,
\[ R, \]
reserves,
and
\[ K, \]
bank capital.
But accounting rules determine how some of these balance-sheet quantities are measured.
One can therefore write bank lending more generally as
\[ L = f \left( r, R, K^{A}, LA^{A}, C^{A}, \mathbb{E} \right), \]
where the superscript \(A\) emphasizes that capital, liquidity, and collateral are accounting or regulatory representations.
The monetary system therefore responds not only to economic fundamentals but to the accounting representation of those fundamentals.
This leads to an important proposition:
\[ \boxed{ \text{Accounting rules can function as implicit parameters of the monetary transmission mechanism.} } \]
They do not replace monetary policy.
But they can amplify, dampen, or delay its effects.
The Accounting Bias framework makes this mechanism explicit.
Suppose the economic value of a banking asset is
\[ E_i, \]
while its accounting value is
\[ A_i. \]
Then
\[ AB_i=A_i-E_i. \]
If
\[ AB_i>0, \]
reported resources exceed the chosen economic benchmark.
If those reported resources enter lending or liquidity constraints, positive Accounting Bias may support additional balance-sheet expansion.
Conversely, if
\[ AB_i<0, \]
the accounting representation may constrain lending relative to the economic position implied by the benchmark.
This suggests a monetary transmission equation:
\[ \Delta M = g(AB_1,AB_2,\ldots,AB_n,Z), \]
where \(Z\) represents interest rates, regulation, credit demand, risk preferences, and other determinants.
This does not imply that Accounting Bias alone determines money.
It proposes that accounting representation can enter the monetary transmission mechanism.
The relationship should not be overstated.
An accounting rule that increases reported assets does not necessarily increase lending.
Banks may already possess excess capital or liquidity. Credit demand may be weak. Risk expectations may deteriorate. Regulatory constraints may dominate accounting quantities. Central-bank policy may offset the effect. Banks may voluntarily choose not to expand their balance sheets.
Therefore,
\[ \Delta A>0 \]
does not imply mechanically
\[ \Delta M>0. \]
A more accurate statement is:
\[ \boxed{ \text{Accounting affects money when accounting quantities enter binding financial constraints or behavioral decisions.} } \]
This is the condition under which the proposed channel becomes economically significant.
The theory generates several testable hypotheses.
First, accounting changes that increase recognized bank capital or liquid assets should have stronger effects on lending among institutions close to capital or liquidity constraints.
Formally,
\[ \frac{\partial L}{\partial A} \]
should be larger when the relevant constraint is binding.
Second, accounting losses should have stronger contractionary effects for highly leveraged institutions.
Third, fair-value changes should matter more for institutions holding larger portfolios of assets subject to market-based remeasurement.
Fourth, impairment rules should affect the timing of credit contraction.
Fifth, collateral revaluations should affect lending most strongly where secured lending dominates.
These predictions turn the theoretical framework into an empirical research program.
The argument also reveals something more general.
Money is not merely a quantity of tokens.
In modern economies, money is embedded in a hierarchy of financial claims.
At one extreme lie central-bank reserves and currency.
Then come commercial-bank deposits.
Then highly liquid short-term financial claims.
Further away lie securities, loans, collateral assets, and productive capital.
The boundary between
\[ \text{money} \]
and
\[ \text{non-money} \]
is therefore partly institutional.
Accounting participates in constructing that institutional hierarchy.
Recognition determines which claims appear. Measurement determines their reported amount. Classification determines their apparent liquidity. Timing determines when losses and gains enter the balance sheet.
The monetary system is therefore partly shaped by the accounting architecture through which financial claims are represented.
Accounting does not create money in the same direct sense that a commercial bank creates a deposit by extending a loan.
But accounting practices can influence the monetary system because money creation takes place through balance sheets.
The key mechanism is
\[ \boxed{ \text{Accounting rule} \rightarrow \text{reported financial position} \rightarrow \text{balance-sheet constraint} \rightarrow \text{credit creation} \rightarrow \text{deposit creation} \rightarrow \text{money supply} } \]
Recognition rules may change which financial resources appear on balance sheets.
Measurement rules may alter reported asset values and equity.
Impairment rules may accelerate or delay losses.
Liquidity classification may affect perceived financial capacity.
Collateral accounting may change borrowing limits.
All of these channels can influence credit creation when the corresponding financial constraints are binding.
The Accounting Bias framework therefore extends naturally into monetary economics.
If
\[ A=\mathcal{A}(E), \]
and monetary creation depends upon reported balance-sheet quantities, then
\[ M = \mathcal{M}[\mathcal{A}(E)]. \]
Accounting is no longer merely a passive record of monetary activity.
It becomes one of the institutional mechanisms through which economic reality is translated into the balance-sheet quantities that govern financial intermediation.
The most important conclusion is therefore not that accounting rules mechanically determine the money supply.
It is more precise:
\[ \boxed{ \textbf{In a balance-sheet monetary system, the rules used to measure the balance sheet can influence the capacity of the system to create money.} } \]