Fractional reserve banking works by letting commercial banks hold only a small portion of customer deposits as cash on hand and lend the rest out, which turns a single deposit into a chain of new loans and new deposits throughout the economy. When you put $10,000 into a checking account, the bank doesn’t lock it in a vault. It lends most of it to someone else. That borrower spends the money, the recipient deposits it at another bank, and the cycle repeats. The result is that the banking system creates most of the money circulating in the economy, and your checking balance is really a claim on the bank rather than a stack of bills sitting somewhere with your name on it.
The Lending Cycle, Step by Step
Start with your $10,000 deposit. Your bank records a $10,000 balance in your account but doesn’t need to hold all of it. Suppose it keeps 10% as an internal buffer. It sets aside $1,000 and lends the remaining $9,000 to a borrower who uses it to buy equipment from a supplier. The supplier deposits that $9,000 at a different bank. That bank now treats it as a brand-new deposit, holds $900 as its own buffer, and lends out $8,100.
Each round shrinks the lendable amount, but each round also creates a new deposit that didn’t exist before. After dozens of iterations, that original $10,000 has generated tens of thousands of dollars in total deposits across the banking system. Your $10,000 still shows in your account. The borrower’s $9,000 loan became real spending power. The supplier’s $9,000 deposit is simultaneously available to them. All of these balances function as money, even though the original cash only entered the system once.
The Money Multiplier
Economists summarize this expansion with a simple formula called the money multiplier: one divided by the reserve ratio. If banks hold 10% of deposits as reserves, the multiplier is 10, meaning each new dollar of deposits could theoretically become $10 of total money. The Federal Reserve tracks this expansion through two measures. M1 includes currency in circulation plus checking account balances. M2 adds savings deposits, small time deposits, and retail money market fund shares on top of M1.
In practice, the multiplier rarely hits its theoretical ceiling. Banks often hold more than the minimum because lending carries risk, and not every dollar lent out gets re-deposited into the banking system. Some borrowers hold cash. Some send payments overseas. Some repay loans faster than new ones are issued. Market conditions also shape how aggressively banks lend. When the economy feels shaky, banks tighten standards and the multiplier weakens. When confidence is high, lending expands.
Since 2008, the connection between reserves and lending has loosened further. The Federal Reserve began paying interest on reserves that year, which gave banks a reason to park excess funds at the Fed instead of lending them out. The St. Louis Fed has noted that this shift largely explains why the textbook money multiplier stopped predicting real-world money creation accurately. Reserves piled up, but lending did not expand proportionally.
Reserve Requirements Are Now Zero
The statutory authority for reserve requirements comes from 12 U.S.C. ยง 461, which lets the Federal Reserve Board require depository institutions to hold reserves against transaction accounts for monetary policy purposes. The Fed exercises this authority through Regulation D (12 C.F.R. Part 204), which spells out the specific ratios and thresholds.
For decades, the system worked in tiers. Smaller banks with modest deposit bases owed little or nothing. Larger institutions with net transaction accounts above a certain threshold were subject to a 10% reserve requirement, a ratio that had been in place since 1992 when it was lowered from 12%. That 10% figure is still the one most textbooks reference.
That changed on March 26, 2020, when the Board of Governors dropped all reserve requirement ratios to 0% for every depository institution. The 3% rate on the lower tier and the 10% rate on the upper tier both went to zero simultaneously. As of 2026, the reserve requirement ratios in Regulation D’s table remain at 0% across all categories, including nonpersonal time deposits and Eurocurrency liabilities.
Banks didn’t empty their vaults as a result. They still hold reserves voluntarily because they need cash to process daily withdrawals, settle transactions, and manage internal risk policies. But the legal floor is gone. What actually constrains lending today is a different set of rules.
What Actually Limits Lending Now
With reserve requirements at zero, capital requirements have become the primary brake on how much a bank can lend. These rules focus not on how much cash a bank keeps in reserve, but on how much of its own money (shareholder equity and retained earnings) it has relative to the risk it takes on.
The core measure is the Common Equity Tier 1 (CET1) capital ratio. Every large bank must hold CET1 capital equal to at least 4.5% of its risk-weighted assets. On top of that, the Fed adds a stress capital buffer of at least 2.5%, determined by annual stress tests that simulate severe economic downturns. The largest globally significant banks face an additional surcharge of at least 1%. A major bank typically needs to hold somewhere between 7% and 10% or more of its risk-weighted assets in high-quality capital.
The phrase “risk-weighted” matters. Not all bank assets count equally. U.S. Treasury securities carry a 0% risk weight, meaning they don’t count against the capital requirement at all. State and local government bonds carry 20% to 50% depending on the type. A standard commercial loan carries 100%. High-volatility commercial real estate loans carry 150%. A bank stuffed with Treasuries can hold far less capital than one heavy on commercial lending, which directly shapes what kinds of loans banks prefer to make.
Every U.S. bank must also meet a simple leverage ratio: Tier 1 capital divided by total balance sheet assets, with no risk weighting. The minimum is 4%, and a bank needs 5% to be classified as “well-capitalized.” For the largest systemically important institutions, a supplementary leverage ratio rule effective April 1, 2026 caps the enhanced standard at 4% for their depository subsidiaries.
What This Means for the Money in Your Account
A bank’s balance sheet shows how fractional reserve banking works in accounting terms. Your deposit is the bank’s liability because it owes you that money on demand. The loan the bank made with your deposit is the bank’s asset because the borrower owes the bank principal plus interest. Reserves, whether sitting in the vault as cash or held as a digital balance at a regional Federal Reserve Bank, are also assets.
This is why banking can feel counterintuitive. The money in your checking account isn’t really “there” in the way most people imagine. It’s a claim against the bank. The bank has already sent most of it out the door as loans and investments, keeping only enough reserves to handle normal daily flows. What backs your balance is the bank’s overall solvency: the quality of its loan portfolio, the value of its securities holdings, and its capital cushion.
When a bank earns more interest on its loans and investments than it pays depositors, the difference is net interest income, which is the primary revenue source for most commercial banks. That spread is the financial incentive that makes the entire fractional reserve system run. Banks want to lend because lending is how they make money.
Bank Runs and the Safety Net
Fractional reserve banking rests on a bet that most depositors won’t demand their money at the same time. When that bet fails, the result is a bank run, and it can destroy an otherwise solvent institution within hours.
Silicon Valley Bank in March 2023 is the most vivid recent example. SVB had invested heavily in long-term securities that lost value as interest rates rose, creating large unrealized losses. When management tried to raise capital and the news spread, depositors, many of whom held balances far above the insurance limit, pulled $40 billion in a single day, with an additional $100 billion in withdrawal requests the bank could not meet. SVB failed within 48 hours. In the era of online banking, a run no longer requires a line around the block.
The primary defense against runs is federal deposit insurance. The FDIC insures deposits up to $250,000 per depositor, per bank, per ownership category. Since the FDIC was founded in 1933, no depositor has lost a penny of insured funds. The Deposit Insurance Fund backing these guarantees carries the full faith and credit of the United States government. When a bank does fail, the FDIC usually makes insured deposits available by the next business day.
Deposit insurance works as much psychologically as financially. If you know your balance is covered, you have no reason to rush to the bank at the first sign of trouble, which means the trouble is less likely to spiral. SVB’s vulnerability was concentrated in large corporate accounts that blew past the $250,000 limit, leaving those depositors exposed and highly motivated to run. For most individuals with balances under the insurance cap, fractional reserve banking functions invisibly. The underlying risk is real, but the safety net, combined with capital rules and Federal Reserve liquidity tools, keeps the system workable for ordinary depositors.