A topic in the Open Knowledge Graph — a free, open map of 15,290 topics and the order to learn them in.
The monetary base (currency plus bank reserves) is what the central bank controls directly through open market operations. Broader money supply (M1, M2) is larger because of deposit multiplication: when banks receive deposits, they lend out a fraction, creating new deposits elsewhere. Money multiplier amplifies base injections.
Trace a deposit through banking system: central bank bond purchase injects base. Bank lends out (1 − reserve ratio) × deposit, creating new deposits. Process repeats, creating money supply equal to base × (1 / reserve ratio). Show why multiplier is lower with higher reserve requirements.
You already know how money creation works through the banking system: banks take deposits and lend out most of them, and those loans become deposits elsewhere, which are lent out again, creating a chain of new money. The monetary base (sometimes called "high-powered money" or M0) is the foundation of this process — it is the currency in circulation plus the reserves that banks hold at the central bank. The monetary base is what the central bank controls directly, primarily through open market operations.
When the central bank buys a government bond from a bank, it credits the bank's reserve account — the monetary base increases by the bond's purchase price. Those new reserves are the raw material for deposit creation. Suppose the required reserve ratio is 10%. A bank receiving $1,000 of new reserves lends out $900, which flows to another bank as a deposit. That bank holds $90 in reserves and lends out $810, which flows to yet another bank — and so on. The total deposit creation eventually sums to $1,000 / 0.10 = $10,000. This is the money multiplier: 1 / reserve ratio. The broader money supply (M1, M2) equals the monetary base times the multiplier.
The mechanics are clean, but the real-world multiplier is far less stable than the textbook formula suggests. Three factors drive a wedge between the theoretical and actual multiplier. First, banks hold excess reserves — reserves beyond the required minimum — when the opportunity cost of holding reserves is low (when interest rates are near zero) or when lending prospects look risky. If banks hold excess reserves instead of lending them out, the multiplication chain is shorter. Second, households and firms hold some fraction of money as cash (currency outside banks) rather than as bank deposits. Cash held in wallets doesn't circulate back as deposits, damping the multiplication process. Third, the required reserve ratio itself is often not the binding constraint: capital requirements, regulatory risk weights, and the availability of creditworthy borrowers all limit lending independently of reserve levels.
The 2008 financial crisis provided a dramatic natural experiment. The Federal Reserve expanded the monetary base from roughly $850 billion to over $4 trillion through quantitative easing programs between 2008 and 2014. Yet M2 grew comparatively modestly, and inflation remained subdued — far below what a naive multiplier model would predict. The reason: banks absorbed the massive base injection as excess reserves (at the time, the Fed paid interest on reserves, making it attractive to park cash at the Fed rather than lend). The predicted multiplication simply did not occur because banks chose not to lend at the scale the theory implied.
This history recalibrates how to think about monetary policy transmission. The central bank controls the price of base money (through interest rates on reserves and the federal funds rate) more reliably than it controls the quantity of broad money. Monetary policy works primarily by changing incentives — for banks to lend, for firms to invest, for households to borrow — rather than by mechanically injecting a fixed multiple into the economy. The money multiplier remains a useful conceptual tool for understanding why central bank balance sheet expansions can potentially be inflationary, but it should be treated as a rough guide to the direction of effects rather than a precise quantitative relationship.