Inside a Single Bank: What Actually Changes When Money Is Created#

From the outside, the banking system multiplies money through an elegant cascade — each bank lending, each borrower spending, each recipient depositing, round after round. From inside a single bank, none of that is visible. There’s no cascade. There’s no multiplier. There’s just a balance sheet, a set of rules, and a series of decisions that, taken individually, seem almost disappointingly ordinary.

The Deposit Arrives#

Consider a mid-sized commercial bank — call it First National. On a Tuesday morning, a local manufacturer deposits $1,000 in cash, proceeds from a recent sale. The moment the teller processes the transaction, two things change on First National’s balance sheet simultaneously.

On the asset side, cash reserves increase by $1,000. On the liability side, the bank records a new demand deposit of $1,000 — money it owes the manufacturer on demand. The bank is neither richer nor poorer. Assets and liabilities grew by equal amounts. Think of a water tank: $1,000 just flowed in through the intake pipe.

The balance sheet — a financial statement showing what a bank owns (assets) and what it owes (liabilities) — must always balance. Every asset has a corresponding liability or equity claim. This isn’t a regulation. It’s an accounting identity, as fundamental as 2 + 2 = 4.

First National’s Balance Sheet: After Deposit#

Total change: +$1,000 on each side. Perfectly balanced.

Calculating the Constraint#

First National operates under a 10% reserve requirement. Of the $1,000 in new deposits, the bank must hold at least $100 as required reserves — funds that can’t be lent, invested, or deployed in any way. They sit in the bank’s account at the central bank or in the vault, serving as a buffer.

The remaining $900 is excess reserves — funds above and beyond what regulation demands. Excess reserves earn the bank little or no return. They represent idle capacity: money that could be generating interest income through loans but currently sits dormant.

The economic calculation is simple. Holding $900 in excess reserves produces negligible income. Lending $900 to a creditworthy borrower at 6% annual interest produces $54 per year. The pressure to lend is intensifying: in April 2026, Customers Bank signed a deal with OpenAI to integrate artificial intelligence into its lending workflow, slashing approval times from 30–45 days to just seven (CNBC). For a bank like First National, such technology would mean excess reserves spend far less time sitting idle — the gap between receiving a deposit and issuing a loan shrinks from weeks to days. The incentive to lend isn’t subtle. It’s the fundamental business model of commercial banking: borrow short (deposits), lend long (loans), pocket the spread.

The Lending Decision#

First National’s loan officer reviews an application from a local restaurant owner seeking $900 for kitchen renovations. The borrower has strong revenue, manageable debt, and equipment as collateral. Loan approved.

Here’s where a critical distinction emerges between how most people picture lending and how it actually works. The common mental model: the bank reaches into its vault, pulls out $900 in cash, and hands it over. That model is wrong.

What actually happens is simpler and more profound. The bank credits the borrower’s checking account with $900. It creates a new deposit by entering numbers into a computer. The borrower’s balance goes up by $900. Simultaneously, the bank records a new asset: a loan receivable of $900 — the borrower’s promise to repay with interest.

First National’s Balance Sheet: After Loan#

Look at what happened. Total deposits at First National jumped from $1,000 to $1,900. The original depositor still has $1,000. The borrower now has $900. The bank created $900 in new deposits — new money — at the stroke of a key. Reserves didn’t change. The balance sheet expanded on both sides simultaneously.

This is the moment of money creation. Not when the central bank prints currency. Not when gold gets mined. Right here, in a routine transaction between a bank and a borrower, new purchasing power enters the economy.

The Water Tank Analogy#

A physical metaphor makes the mechanics clearer. Picture First National as a water tank with three compartments.

The first compartment holds required reserves — $100 of water sealed off by regulation. It cannot flow out under any circumstances.

The second holds excess reserves — $900 of water that can flow out through the lending pipe. This is the bank’s capacity to create new money.

The third represents loans outstanding — water that has already flowed out to borrowers. It no longer sits in the tank, but the bank holds a claim on it (the loan contract), expecting it to flow back over time with interest.

When a deposit arrives, water flows in. The required-reserve compartment fills first. The rest accumulates as excess reserves. When a loan is issued, water flows from excess reserves out through the lending pipe. Total water level (assets) stays the same, but the composition shifts: less liquid reserves, more illiquid loans.

The tank can’t lend water it doesn’t have. It can’t drain the required-reserve compartment. These physical constraints mirror the financial rules governing real banks. The analogy has one important limitation — banks create deposits at the moment of lending rather than transferring existing funds — but it captures the essential dynamics of reserve management.

The Money Flows Out#

The restaurant owner didn’t borrow $900 to admire a checking account balance. Within days, the borrower writes checks and makes payments: $400 to a contractor, $300 to an equipment supplier, $200 to a food distributor. All three bank elsewhere.

As these payments clear, $900 in reserves flows out of First National to the recipient banks. The borrower’s deposit drops to zero (spent in full). But the loan receivable stays — the borrower still owes $900 plus interest.

First National’s Balance Sheet: After Funds Transfer#

First National is back to its original size — $1,000 in assets, $1,000 in liabilities. But the composition transformed. Instead of $1,000 in liquid reserves, the bank now holds $100 in reserves (the required minimum) and $900 in loans. Idle liquidity became an income-producing asset.

The original depositor’s $1,000 remains fully available. The bank can honor a withdrawal because the reserve requirement is designed to ensure sufficient liquidity for normal operations. The system works as long as all depositors don’t demand their money simultaneously — the scenario known as a bank run.

What First National Cannot Do#

The balance sheet tells a story of limits as much as possibilities. First National received $1,000 and lent $900. It did not lend $9,000 or $10,000. It did not “multiply” anything. It lent its excess reserves — the amount above what regulation required it to hold.

This deserves emphasis because it contradicts a persistent myth. Individual banks don’t multiply deposits. They lend a fraction of them. The multiplication happens across the system, as the $900 lent by First National becomes a deposit at Bank B, which keeps $90 and lends $810, which becomes a deposit at Bank C, and so on.

First National is one link in a chain. It forged its link by performing a single, bounded operation: retain reserves, lend the rest. The chain’s total length — the system-wide multiplier — comes not from any single bank’s action but from the cumulative effect of every bank performing the same operation in sequence.

The Balance Sheet as a Window#

Looking at a single bank’s balance sheet reveals something the system-level view obscures: the profound gap between what a bank does and what the system achieves. First National’s loan officer didn’t think in terms of money multipliers or deposit expansion cascades. The officer evaluated a borrower, assessed risk, and approved a loan. The macro consequence — contributing to system-wide money creation — was completely invisible at the point of decision.

This asymmetry isn’t unique to banking. A single ant doesn’t grasp the architecture of the colony. A single neuron doesn’t comprehend the thought it helps produce. Complex system-level behaviors frequently emerge from simple, local actions by individual agents following simple rules. Banking works the same way.

The Credit-Creation model of banking, viewed from the micro level, is a story about individual banks making individual decisions within individual constraints. Each bank sees only its own balance sheet, its own reserve position, its own loan portfolio. The systemic multiplication is an emergent property — a pattern that exists at the level of the whole but is invisible at the level of any single part.

The Tip of the Iceberg#

First National’s experience is one visible fragment of a much larger process. Below the surface, hundreds or thousands of banks simultaneously receive deposits, calculate reserves, evaluate borrowers, and issue loans. Each performs the same basic operation. Each contributes a diminishing increment to the total money supply.

The micro view also exposes vulnerabilities invisible from above. If First National’s borrower defaults on the $900 loan, the bank absorbs the loss against its capital. If many borrowers default at once, capital erodes, lending capacity contracts, and the bank’s contribution to the money creation machinery stalls. Scale that across many banks, and the system-wide multiplier collapses — not because the math changed, but because the individual actors at the base of the cascade stopped functioning.

The 2008 financial crisis played out exactly along these lines. Banks facing mounting loan losses pulled back from lending. Deposit expansion slowed and, in some sectors, reversed. Deleveraging — reducing outstanding loans — destroyed deposits just as effectively as lending had created them. Money creation runs in both directions.

From One Bank to the Whole System#

The trip inside First National’s balance sheet confirms a counterintuitive truth: the power of money creation doesn’t live in any single institution. A bank is a valve, not a pump. It regulates the flow of credit according to its reserves, its capital, and the demand it faces. The pressure that drives the flow comes from the system — from the continuous circulation of deposits and loans across thousands of institutions.

Understanding what happens inside one bank makes the system-level view more concrete and less mysterious. The multiplier isn’t an abstract formula imposed from above. It’s the cumulative result of thousands of banks, each running the same simple calculation: how much must I hold, and how much can I lend?

The next question follows naturally. If individual banks are merely links in a chain, what determines the chain’s behavior as a whole? What happens when links interact, when one bank’s lending shifts another’s reserves, when confidence rises or falls across the system at once? The answer lies in the emergent properties of the banking system — patterns that appear only when the whole turns out to be greater, and sometimes more dangerous, than the sum of its parts.