Module 01 established what a demand deposit is and why people accept a private institution's promise as money. Now we open the machine. A bank is, at bottom, a balance sheet — a particular arrangement of what it owns and what it owes — and almost everything that makes banking powerful, profitable, and dangerous can be read directly off that balance sheet. This module covers reserves, the money multiplier and the more accurate "loans create deposits" view, maturity transformation, the spread that pays for it all, and the thin sliver of capital standing between a bank and failure.
The most useful single idea in this module is that a bank is not really a building or a brand — it is a balance sheet. Everything it does shows up as a change to what it owns (its assets) or what it owes (its liabilities), and the gap between the two is the owners' stake (its equity or capital). Learn to read the balance sheet and banking stops being mysterious.
The first thing that surprises newcomers is where the deposits go. From Module 01 you already know a deposit is a claim you hold against the bank — a promise the bank owes you. So on the bank's books, your deposit is a liability, not an asset. It is money the bank owes back to you. The money the bank lends out, by contrast, is an asset — a promise borrowers owe to the bank. This inversion trips everyone up at first: the money in "your" account is something the bank owes, and the loans the bank made are something it owns.
Here is a stylized bank balance sheet, expressed per $100 of total assets so it doesn't pretend to be any specific institution:
| Assets (what it owns) | $ | Liabilities & equity (what it owes) | $ |
|---|---|---|---|
| Reserves (cash + central-bank balances) | 6 | Deposits | 88 |
| Loans | 80 | Other borrowings | 4 |
| Securities | 14 | Equity (capital) | 8 |
| Total assets | 100 | Total liabilities & equity | 100 |
Three features of this picture drive the entire rest of the module: reserves are a small fraction of deposits (here 6 against 88), the bulk of the assets are loans and securities that cannot be turned into cash instantly, and equity is a thin sliver (8) relative to the whole. Each of those is a deliberate choice, and each carries a consequence we will trace in turn.
Deposits are the bank's liabilities; loans are its assets. The money you think of as "yours, held by the bank" is money the bank owes you and has mostly lent to someone else. The whole craft of banking is managing the mismatch between liabilities that can be called instantly and assets that cannot.
Reserves are the truly liquid part of a bank's assets: physical cash in its vaults and, more importantly, the balances it holds at the central bank. Reserves are what a bank actually uses to settle withdrawals and to pay other banks. When you take cash from an ATM or your payment clears to another bank, it is ultimately reserves that move.
The reserve ratio is the fraction of deposits a bank holds as reserves. Hold too little and you cannot meet withdrawals; hold too much and you forgo the interest you could have earned by lending. Historically, governments set a required reserve ratio — a legal minimum — and banks held a little extra as a buffer (excess reserves).
This is a place where the textbook and the modern world have drifted apart, and the international picture makes the point vividly:
China treats the reserve requirement ratio (RRR) as a major monetary-policy tool, raising and lowering it deliberately to tighten or loosen credit — a lever it pulls several times a year. Canada and the United Kingdom have operated with no mandatory reserve requirement for decades, relying on other tools. The United States required reserves for a century but cut the requirement to zero in March 2020, shifting to a system where banks hold ample reserves voluntarily and the central bank steers rates by paying interest on them.
The lesson is not which number is "right." It is that the required reserve ratio — once presented as the central dial of banking — is in much of the world either a deliberate policy instrument or simply absent. What never disappears is the underlying tension: a bank must keep enough liquid reserves to meet whatever withdrawals actually arrive, whether or not a law forces it to.
The classic story of how banking expands the money supply goes like this. Suppose the reserve ratio is 10%. Someone deposits $1,000. The bank keeps $100 as reserves and lends out $900. That $900 gets spent and lands in another bank as a deposit; that bank keeps $90 and lends $810. The $810 becomes a deposit elsewhere, and so on. Add up the whole chain and the original $1,000 of reserves supports up to $10,000 of deposits — the original deposit times one divided by the reserve ratio.
| Round | New deposit | Kept as reserve (10%) | Lent on |
|---|---|---|---|
| 1 | 1,000 | 100 | 900 |
| 2 | 900 | 90 | 810 |
| 3 | 810 | 81 | 729 |
| … | … | … | … |
| Total | 10,000 | 1,000 | 9,000 |
This is the money multiplier, and it captures something real: a banking system with fractional reserves creates far more deposit money than the cash that seeded it. The figure 1 ÷ reserve ratio is the theoretical maximum expansion. It is a clean, memorable model, and it is the version most people first learn.
It is also, as a description of how banks actually behave day to day, somewhat backwards — which is the subject of the next section.
In a modern economy with a central bank, banks do not wait for a deposit before lending. When a creditworthy borrower walks in and the loan looks profitable, the bank makes the loan first — and, as Module 01 noted, it does so by simply crediting the borrower's account, creating a new deposit out of nothing. Only afterward does the bank worry about obtaining whatever reserves it needs to settle, and in a modern system it can borrow those reserves from other banks or from the central bank at the prevailing interest rate.
So the causation in the textbook multiplier runs the wrong way. It is not "reserves come in, then a fraction is lent." It is closer to "a loan is made, which creates a deposit, and reserves are obtained as needed." This is not a heterodox opinion; it is how major central banks now describe their own systems. The Bank of England's widely cited 2014 explanation put it plainly: in the real world, lending creates deposits, not the other way around, and the central bank supplies reserves to support the resulting payments.
What, then, actually limits how much banks lend, if not a pile of reserves waiting to be multiplied? Three things, none of them the reserve ratio:
The multiplier is a fine way to see that a fractional-reserve system supports far more deposits than base cash. But the operative reality is that banks create money by lending, and reserves follow. Both views agree on the punchline from Module 01 — most money is bank-created — but the modern view correctly locates the constraint on money creation in profitable demand and bank capital, not in a fixed multiple of reserves.
Look again at the balance sheet in Section 01. The liabilities (deposits) can be demanded back instantly. The assets (loans) repay over years — a mortgage over thirty, a business loan over five. The bank has deliberately funded long-dated, illiquid assets with short-dated, instantly-callable liabilities. This is maturity transformation, and it is not a side effect of banking; it is the core economic service banks provide.
Why is it valuable? Because savers want their money available at a moment's notice, while the economy's most productive uses of money — building a factory, buying a home, funding a company — need money committed for years. Maturity transformation bridges that gap. The bank pools many depositors, relies on the statistical fact that they will not all withdraw at once, and channels their collectively-stable balances into long-term lending. Society gets long-term investment funded by short-term savings. That is a genuine economic good, and it is why banks earn their keep.
The cost of providing it is the vulnerability built into the structure. Because the long assets cannot be liquidated quickly, a bank that suddenly faces more withdrawals than its reserves can cover is in trouble even if every one of its loans is sound — the money is simply tied up. Maturity transformation creates value in normal times and the risk of collapse in abnormal ones. The same single feature is the source of both. Module 04 is devoted to what happens when it breaks.
If a bank owes its depositors money and owns loans to borrowers, where does its profit come from? Mostly from the spread between the two: it pays depositors a low rate (often near zero on checking balances) and charges borrowers a higher rate, pocketing the difference. This difference, scaled across the whole balance sheet, is the net interest margin — the engine of traditional bank profitability.
A simplified illustration: if a bank pays an average of 1% on its funding and earns an average of 5% on its loans and securities, its net interest margin is roughly 4 percentage points on its assets. On $100 of assets that is about $4 of gross interest income a year, out of which it must cover salaries, branches, technology, loan losses, and a return to shareholders. Banks supplement this with fee income — account fees, card interchange, payment charges, advisory fees — which has grown more important as interest margins have compressed in many markets.
Notice the incentive this creates, because it sets up the next module. The spread is earned only on assets that pay interest — overwhelmingly, loans. A bank that merely warehoused deposits and held safe reserves would earn almost nothing and could not cover its costs. The business model therefore pushes every deposit-taking bank toward lending. This is the crucial hinge: lending is not required in order to take deposits and run a payment system — but it is how the conventional bank makes money, so conventional banks nearly all do it. Module 03 takes up the consequences of that near-universal choice, and Module 06 returns to whether an institution could profitably not lend.
Net interest margin is earned on loans, not on reserves. Deposit-taking and payments could in principle be offered by an institution that never lends — but it would earn almost nothing under the spread model and would have to charge customers directly. So conventional banking bundles storage, payments, and lending together not out of necessity but out of profitability. Keeping that distinction in view is what lets you see the later innovations clearly.
Return one last time to the balance sheet. Equity was just 8 out of 100. That sliver is the bank's capital — the owners' own money at stake — and it is what absorbs losses. If borrowers default and the loan book falls in value by 5, that loss eats into the 8 of equity, leaving 3. Depositors are untouched, because the owners' stake took the hit first. Capital is the shock absorber that stands between bad loans and broken deposit promises.
Because equity is small relative to assets, banks are highly leveraged. An 8-to-100 structure means the bank holds assets worth about twelve times its own capital. That leverage is what magnifies returns in good times — a small spread on a large balance sheet becomes a healthy return on a thin equity base — and what makes banks dangerous in bad times, because it does not take a large percentage loss on assets to wipe out the whole equity cushion. A 100%-equity-funded business cannot fail from a modest loss; a 12-times-levered one can.
This is why bank capital is regulated so heavily. The global framework, known as Basel (after the city where the standard-setting committee meets), requires banks to hold minimum capital scaled to how risky their assets are, plus a simple leverage ratio backstop — a floor on equity relative to total assets regardless of risk weighting, set at a minimum of 3% under Basel III. We treat the safety net and its rules properly in Module 05; for now, just hold the idea that capital is the loss-absorbing buffer and that banks deliberately keep it thin.
Put the pieces together and the bank becomes a coherent — and precarious — machine. It funds itself mostly with deposits it owes on demand, holds only a small slice of liquid reserves against them, deploys the rest into loans and securities that repay slowly, earns the spread between what it pays and what it charges, and rests the whole structure on a thin layer of its own capital. Each design choice serves a purpose, and each creates an exposure:
None of these is an accident or an abuse. They are the normal, designed condition of a conventional bank, and together they explain why banking is simultaneously one of the most useful and one of the most failure-prone institutions in any economy. They also explain why the state gets so deeply involved — the subject of Module 05 — and why a balance sheet this fragile is something newcomers are mostly not allowed to operate, which is the near-monopoly Module 06 examines.
The immediate next step, though, is to sit with the most uncomfortable implication of everything above. If your deposit is the bank's liability, and the bank has used it to make loans it owns, then the money you treat as plain cash is in fact a claim on a leveraged lender — and you are, whether you chose to be or not, exposed to that lender's credit. Module 03 confronts this directly and traces the web of credit it creates across the whole banking system.
Six questions on the bank balance sheet and the mechanics of money creation. The questions test the concepts and their consequences rather than memorization of any particular ratio.