What Is Money? Functions, Types and How Modern Money Works

What is money illustration showing coins, banknotes and digital payments connected to money’s core economic functions.

Money is something we use every day, yet defining it is less straightforward than it may seem. It is not limited to banknotes and coins, nor does an asset become money simply because people assign it value or call it a currency.

Economists generally define money by the functions it performs. Money provides a medium of exchange for buying and selling goods and services, a unit of account in which prices and debts can be expressed, and a store of value that allows purchasing power to be carried forward through time.

These functions matter beyond textbook economics. Modern money exists largely in electronic form as deposits held at commercial banks, while central banks use broader statistical measures to distinguish money from other financial assets. Understanding those distinctions is important when examining monetary policy, money creation and exchange rates.

They also provide a useful framework for investors. Foreign exchange involves currencies issued within established monetary systems, whereas an asset described as a “cryptocurrency” does not necessarily perform all the economic functions associated with money. Whether Bitcoin, a stablecoin, gold or another asset should be considered money is therefore a question of function rather than terminology.

Money Is Defined by What It Does

Money can take different forms. Historically, societies have used commodities and precious metals; modern monetary systems rely primarily on fiat currency and commercial-bank deposits. The physical form is secondary to whether people can reliably use the asset to make payments, express prices and transfer purchasing power through time.

This functional definition also explains why the boundary between money and other financial assets is not always perfectly sharp. Some assets perform certain monetary functions well but others poorly. Central banks therefore measure the money supply using different monetary aggregates, ranging from relatively narrow definitions of highly liquid money to broader measures that include additional liquid financial assets.

Before examining those distinctions, it helps to look more closely at the three core functions that make money useful.

The Functions of Money

Economists usually describe money through three core functions: it acts as a medium of exchange, a unit of account and a store of value. These functions reinforce one another, but each solves a different economic problem.

Medium of Exchange

A medium of exchange allows people to buy and sell without directly swapping one good or service for another. Economists often illustrate this benefit with a hypothetical barter economy.

Suppose a farmer wants a haircut, but the barber has no use for vegetables. A direct trade cannot occur unless both parties want what the other offers at the same time. Economists call this the double coincidence of wants.

Money removes that constraint. The farmer can sell produce for money and use the proceeds to pay the barber. The barber can then spend the same money elsewhere. This works because people generally expect others within the same monetary system to accept it in payment.

The barter example is a useful economic model. It does not prove that societies historically progressed from pure barter to money in one simple sequence.

Unit of Account

Money also provides a common unit for expressing prices, debts and financial values.

Without one, people would have to compare the value of goods against many other goods. A laptop might need a price in barrels of oil, hours of labour, kilograms of wheat and countless other items.

A common monetary unit removes that complexity. It makes prices easier to compare and provides a standard for accounting and contracts. Businesses and households can express wages, assets, liabilities and future payments in the same unit.

This function does not require physical cash to change hands. A bank deposit or financial contract can still use dollars, euros or another currency as its unit of account.

Store of Value

Money can also transfer purchasing power from the present to the future. Someone who receives income today can hold part of it and spend it later.

That does not mean money preserves wealth perfectly. Inflation reduces a currency’s purchasing power over time, sometimes substantially. Money can therefore serve as a store of value without being an ideal long-term investment.

This distinction will become important when we compare conventional currencies with other assets. Something may hold or increase its market value while functioning poorly as a medium of exchange or unit of account.

Is Standard of Deferred Payment a Fourth Function?

Many introductory explanations focus on the three functions above. Some formal frameworks, including the IMF’s monetary-statistics framework, distinguish a fourth: standard of deferred payment.

This function allows people to express obligations that span time in a common monetary unit. Loans, bonds and invoices, for example, can specify payments that fall due in the future.

The concept overlaps with money’s roles as a unit of account and medium of exchange. For that reason, many economists continue to describe money in terms of three core functions rather than treating deferred payment as a completely separate category.

What Makes Something Useful as Money?

Not every valuable asset works well as money. To perform the functions described above, an asset also needs practical characteristics that make people willing and able to use it.

What Makes Something Useful as Money?

Select a characteristic to see why it matters. Good money combines practical usability with confidence that others will continue to accept it.

Money Design Lab
Characteristic Acceptability
FINANSIFIED · finansified.com

These characteristics help explain why some assets perform monetary functions better than others. Gold, for example, is durable, scarce and divisible. However, it is less convenient for everyday payments than modern bank deposits. A highly volatile financial asset may be easy to transfer and divide, yet still perform poorly as a unit of account.

No characteristic works in isolation. Money depends on a combination of usability, acceptance and confidence that others will continue to recognize its value.

This distinction becomes especially important when evaluating newer forms of payment and digital assets. An asset can be scarce, portable and actively traded without functioning particularly well as money.

How Money Changed Form

Money has taken many forms across different societies and periods. There was no single universal path from barter to coins and then to paper money. What mattered was whether a particular system could perform the functions of money effectively.

Commodity Money

Some societies used commodities that had value beyond their monetary role. Precious metals such as gold and silver became prominent examples because they were durable, divisible and relatively scarce. Other goods have also served monetary purposes in different places and periods.

Commodity money carries value partly because the underlying material has uses of its own. A gold coin, for example, contains a commodity that remains valuable even if people stop using the coin for everyday payments.

Representative Money

Holding and transporting valuable commodities can be inconvenient. One solution was to deposit them with a trusted institution and use transferable claims instead.

In seventeenth-century England, goldsmiths accepted deposits of gold and silver and issued receipts or notes. People increasingly used those claims to make payments rather than moving the metal itself. Early Bank of England notes later developed from similar deposit receipts.

Such representative money derives its monetary value from a claim on an underlying asset. A holder can, in principle, redeem it for the commodity or asset that backs it.

Fiat Money

Modern currencies generally operate differently. A dollar, euro or pound is not a claim that entitles its holder to a fixed quantity of gold or another commodity.

Instead, fiat money depends on acceptance, confidence in the monetary system and the institutions that support it. Its purchasing power also depends on the prices of goods and services people can buy with it.

This shift matters because modern money no longer requires a scarce physical commodity to back every monetary unit. It also means that understanding money today requires looking beyond banknotes and coins. Most money in modern economies exists as balances recorded within the banking system.

What Forms Does Modern Money Take?

Modern money is much more than notes and coins. In most advanced economies, people make the majority of payments by transferring balances held at commercial banks.

It helps to distinguish three closely related forms: physical currency, commercial-bank deposits and central-bank reserves.

Cash

Banknotes and coins are the most visible form of money. People can hold and use them directly without maintaining a commercial-bank account.

Modern banknotes are a form of fiat money. Their value does not come from the paper itself or from a promise to redeem them for a fixed amount of gold. They function as money because people accept them at their stated value within the monetary system.

The exact arrangements for issuing banknotes and coins vary between jurisdictions. Central banks normally issue banknotes, while governments or other public authorities may issue coins.

Commercial-Bank Deposits

The balance displayed in a checking or savings account is also money. It represents a liability of the commercial bank to its customer.

For most households and businesses, bank deposits are the dominant form of money used for everyday electronic payments. A card purchase, bank transfer or direct debit usually moves commercial-bank money rather than physical cash.

Crucially, people normally treat a $100 bank deposit as equivalent to $100 in cash. The ability to transfer deposits and convert them into central-bank money at face value helps different forms of money function as part of the same monetary system.

Central-Bank Reserves

Central-bank reserves are another form of money, but households and ordinary businesses generally do not use them directly.

Banks and other eligible institutions hold reserve balances in accounts at the central bank. They use these balances to settle payments with one another and to meet liquidity needs.

Suppose a customer at Bank A sends money to a customer at Bank B. Their deposits change on the two commercial banks’ balance sheets. Behind the scenes, the banks can settle the transaction by transferring central-bank reserves.

This distinction becomes important when discussing money creation. Commercial banks create deposit money, while central banks create central-bank money. These forms interact closely, but they are not created in the same way.

How Commercial Banks Create Money

Most money in modern economies consists of commercial-bank deposits. Banks create much of this money when they make loans.

Suppose a bank approves a $100,000 mortgage. It does not normally take $100,000 from another customer’s deposit account and transfer that money to the borrower. Instead, the bank records two new entries on its balance sheet:

  • a $100,000 loan, which is an asset of the bank; and
  • a $100,000 deposit in the borrower’s account, which is a liability of the bank.

The loan and deposit appear at the same time. In this sense, bank lending creates new deposit money.

When the borrower spends the money, it may move to an account at another bank. The banks then settle the resulting payment between themselves, typically using central-bank reserves.

Why Banks Cannot Create Unlimited Money

The ability to create deposits does not give commercial banks an unlimited supply of money.

Banks must remain profitable and manage credit risk. They also face capital and liquidity requirements, funding costs and other regulatory constraints. They need enough liquid assets, including central-bank reserves, to meet payment outflows and other obligations. Borrower demand and monetary policy also affect how much lending takes place.

This is why the traditional money-multiplier explanation can be misleading when presented as the mechanism of modern bank lending.

In that textbook model, a bank receives a deposit, keeps a fixed percentage as reserves and lends the remainder. The recipient deposits that money elsewhere, allowing another bank to lend another fraction. Repeating the process supposedly expands deposits according to a fixed reserve ratio.

Modern banking does not operate through such a mechanical sequence. Banks can create a deposit when they make a loan, while their need for reserves and funding arises from settlement, liquidity management, regulation and the structure of their balance sheets.

Loan repayment reverses part of the process. When borrowers repay the principal on bank loans, the corresponding deposit money is removed from circulation. Commercial-bank lending can therefore both create and destroy money.

How Central Banks Measure Money

There is no single statistical boundary between money and other financial assets. Cash clearly functions as money, but many deposits and short-term financial instruments can also be converted into spendable funds quickly.

Central banks therefore use monetary aggregates to group assets according to their liquidity and monetary characteristics.

Narrow and Broad Money

Narrow money includes the assets that people can use most readily for payments. Broader measures add increasingly less liquid assets that can still serve as close substitutes for money.

The exact definitions depend on the monetary system.

In the United States, the Federal Reserve currently publishes two main measures:

  • M1 includes currency, demand deposits and other liquid deposits, including savings deposits.
  • M2 includes M1 plus small-denomination time deposits and retail money market fund balances, subject to several statistical adjustments.

The euro area uses three main aggregates:

  • M1 includes currency in circulation and overnight deposits.
  • M2 adds certain short-term deposits.
  • M3 adds instruments such as repurchase agreements, money market fund shares and certain short-term debt securities.

The labels therefore look similar, but their contents are not identical. Researchers and investors should check the underlying definition before comparing monetary aggregates across countries or historical periods.

The Monetary Base Is Different

Another measure, the monetary base, focuses on central-bank money. It generally includes currency in circulation and reserve balances held at the central bank.

That makes the monetary base different from broad money. Commercial-bank deposits can make up a large share of the money used by households and businesses without appearing as central-bank reserves.

This distinction also helps explain why changes in central-bank reserves do not translate mechanically into proportional changes in the broader money supply. Monetary aggregates reflect the interaction between banks, borrowers, depositors, financial markets and the central bank rather than a fixed reserve multiplier.

Is Cryptocurrency Actually Money?

The word currency in cryptocurrency does not automatically make a cryptoasset money in the economic sense. The better question is whether the asset actually performs the functions of money.

That distinction also avoids treating every cryptoasset alike. Bitcoin and other unbacked cryptoassets work differently from stablecoins that aim to track the value of an existing currency.

Does Bitcoin Function as Money?

Bitcoin can transfer value between users, so it can function as a medium of exchange in some circumstances. However, acceptance remains far narrower than for established currencies. Most businesses still price goods, wages and contracts in dollars, euros or other conventional monetary units rather than bitcoin.

This also limits Bitcoin’s role as a unit of account. Even when a merchant accepts bitcoin, the underlying price may still originate in a conventional currency and then convert into bitcoin at the prevailing exchange rate.

Its store-of-value function presents another problem. Bitcoin has sometimes generated large long-term gains, but an asset can appreciate substantially without providing a stable way to carry purchasing power through time. Large and unpredictable price swings make future purchasing power uncertain.

None of this means Bitcoin has no value or economic use. It means that an investable and transferable asset does not necessarily function well as money.

Stablecoins Are a Different Case

Stablecoins attempt to address one of the main limitations of unbacked cryptoassets: price volatility.

A dollar-denominated stablecoin, for example, aims to maintain a value close to one US dollar. Depending on its structure, the issuer may support that value with cash, bank deposits, government securities or other reserve assets.

This design makes some stablecoins more practical as a medium of exchange and short-term store of value. Stablecoins have consequently become an important settlement instrument within crypto markets. Around 98% of stablecoin value is denominated in US dollars, according to 2026 BIS research.

But this creates an important distinction. A dollar stablecoin generally does not establish an independent monetary unit. It borrows its unit of account and intended price stability from the dollar.

Stablecoins also introduce risks that conventional cash does not have. Their reliability depends on factors such as the quality and liquidity of reserve assets, redemption arrangements, governance and confidence in the issuer. Not every stablecoin offers the same protections.

Digital Assets and Digital Money Are Not the Same Thing

The technology used to record or transfer an asset does not determine whether it qualifies as money.

Commercial-bank deposits are already predominantly digital, yet they form part of the conventional monetary system. A cryptoasset can also exist entirely in digital form without functioning as money.

The useful distinction is therefore not physical versus digital or even traditional versus crypto. It is functional.

Does the asset provide a widely usable medium of exchange? Does it serve as a common unit of account? Can it transfer purchasing power through time with reasonable reliability?

Calling something a currency does not answer those questions.

What FX and Crypto Traders Should Understand About the Data

Four Statistical Features That Complicate Financial Models

See how nonstationarity, non-normal distributions, volatility clustering and multicollinearity appear—and why they matter for trading models and forecasts.

Illustrative data
quant_lab / diagnostics
LEVEL_SERIES SIMULATION
Explanation & implications Nonstationarity

01 / TIME SERIES

Nonstationarity

The series wanders instead of fluctuating around a stable level. Its statistical properties can change through time.

LOOK FOR: drifting level · changing mean · persistent shocks
Implications for trading

Relationships and thresholds estimated in one period may not persist, and trending variables can create spurious-looking signals.

Implications for forecasting

Models fitted directly to nonstationary levels can produce unstable inference and forecasts unless the time-series structure is handled appropriately.

TYPICAL RESPONSE: test for unit roots · difference or detrend where appropriate · consider cointegration
FINANSIFIED · finansified.com · simulated illustrations, not market data

Whether an asset functions as money and how its market price behaves are two separate questions.

Bitcoin and a conventional currency differ substantially in economic structure. Yet their traded price series can share several statistical properties. Those properties matter when investors build models, test relationships or estimate risk.

Price Levels Are Often Nonstationary

Financial price levels often change their statistical properties over time. Their average level today may bear little relationship to their average level several years ago.

Econometricians describe many such series as nonstationary.

Nominal exchange rates often behave approximately like unit-root processes in levels. Bitcoin prices also commonly fail stationarity tests in levels.

This creates a serious modelling problem. Regressing one trending price series on another can produce a high R² and statistically significant coefficients even when no meaningful relationship exists. Economists call this a spurious regression.

For that reason, analysts often transform price levels into returns before modelling them. A common transformation is the logarithmic return.

Return series are often much closer to stationary than the underlying price levels.

This does not mean traders should automatically difference every dataset. Stationarity should be tested rather than assumed.

Returns Are Not Normally Distributed

Another common mistake is to assume that financial returns follow a normal distribution.

FX returns exhibit fat tails. Extreme price changes occur more frequently than a normal distribution would predict. Cryptocurrency returns display the same broad characteristic, often with particularly large extremes.

This matters for risk management.

A model based entirely on normally distributed returns can underestimate the probability of unusually large gains or losses. Measures such as standard deviation remain useful, but they do not describe the whole distribution.

Volatility Changes Over Time

Financial markets also exhibit volatility clustering.

Large price changes tend to occur near other large changes. Quiet periods tend to cluster with other quiet periods.

As a result, volatility is not constant through time.

FX markets have displayed this property extensively, and research finds similar behaviour in Bitcoin and Ethereum. Models that assume constant variance can therefore miss an important feature of market risk.

This is one reason financial econometrics often uses models such as ARCH and GARCH when analysing return volatility.

Correlation Is Not Cointegration

Two asset prices can move upward together for long periods and display a high correlation without sharing a stable economic relationship.

For nonstationary variables, analysts often need to ask a different question: are the series cointegrated?

Cointegration occurs when two or more nonstationary series share a long-run relationship such that a particular combination of them is stationary.

Foreign-exchange economics provides a classic example. Researchers have used cointegration tests to examine long-run relationships between exchange rates and relative price levels under purchasing power parity.

Crypto markets can also exhibit cointegrating relationships. However, analysts should test for them rather than infer them from charts or correlations.

Cointegration does not mean two prices must move together every day. Short-run deviations can occur while a longer-run equilibrium relationship remains intact.

Multicollinearity Is a Different Problem

Multicollinearity should not be confused with cointegration.

It occurs when explanatory variables in a regression contain closely overlapping information. For example, several interest-rate measures or highly related macroeconomic indicators may move closely together.

A model can still fit the data well, but individual coefficient estimates may become unstable. Their standard errors can rise, and it becomes harder to identify the contribution of each variable.

Adding more correlated indicators therefore does not necessarily create a better trading model.

Statistical Relationships Need Economic Interpretation

These properties have a practical implication.

A convincing chart or regression is not evidence of a durable trading relationship by itself. Analysts need to ask whether the variables are stationary, whether volatility changes through time, whether extreme observations dominate the results, and whether any apparent long-run relationship survives proper testing.

They should also ask a more basic question:

Why should these variables be related economically?

Statistical tests can help evaluate a hypothesis. They cannot supply the economic mechanism that the hypothesis lacks.

Why Money Matters for FX and Crypto Investors

Foreign exchange is different from trading a single standalone asset. An exchange rate is the relative price of two currencies, each connected to its own monetary system.

That means investors need to think in relative terms. Interest rates, inflation expectations, monetary policy and expected returns can change the attractiveness of assets denominated in one currency relative to another. Capital can then move across borders, creating pressure on exchange rates.

The important variable is often not the level of one country’s interest rate by itself, but how monetary conditions compare with those elsewhere. A change that markets already expected may also have little effect because traders may have incorporated it into prices before the announcement.

Exchange rates therefore reflect more than changes in the supply of money. They respond to expectations, relative monetary conditions, capital flows, risk sentiment and other economic shocks.

Cryptoassets add another layer. Bitcoin may trade against the US dollar and respond to changes in monetary conditions, but it is not another national monetary system comparable to the euro or yen. Research has found that tighter US monetary policy can affect crypto prices through investors’ willingness to take risk, much as it affects other risky financial assets.

This reinforces an important distinction for investors: economic structure and statistical behaviour are not the same thing.

FX and crypto returns can share properties such as fat tails and volatility clustering. That does not mean the economic forces behind their prices are identical. A useful model needs both statistical evidence and a plausible mechanism explaining why the variables should be related.

Conclusion

Money is best understood by what it does rather than by the form it takes. It provides a medium of exchange, a unit of account and a store of value. Modern monetary systems perform those functions through a combination of physical currency, commercial-bank deposits and central-bank money.

That distinction also helps explain how money is created. Commercial banks can create new deposit money when they extend credit, while central banks create central-bank money and influence financial conditions through monetary policy. Modern money creation therefore cannot be reduced to a mechanical reserve multiplier.

The same functional approach is useful when evaluating newer assets. Calling something a cryptocurrency does not automatically make it money. Bitcoin can transfer value and may serve some monetary functions, but its limited use as a unit of account and large price fluctuations distinguish it from established currencies. Stablecoins occupy a different position because they deliberately reference conventional monetary units.

For investors and traders, these economic distinctions matter alongside the statistical properties of market prices. FX and crypto returns may both exhibit fat tails, volatility clustering and other familiar characteristics, yet the mechanisms that drive them can be very different.

Understanding what money is will not tell an investor where EUR/USD or Bitcoin will trade tomorrow. It does provide a better framework for asking why monetary assets have value, how modern money enters the economy and whether the relationships observed in financial data make economic sense.

Frequently Asked Questions

What are the main functions of money?

Money typically performs three core functions: it acts as a medium of exchange, a unit of account and a store of value. Some economic frameworks also distinguish a fourth function, the standard of deferred payment, which describes money's role in contracts and obligations that extend through time.

Is money the same thing as cash?

No. Banknotes and coins are only one form of money. In modern economies, households and businesses hold much of their money as commercial-bank deposits and use those balances for electronic payments. Central-bank reserves form another type of money, although ordinary households generally cannot use them directly.

Do banks simply lend out money that customers deposit?

Not in the simple textbook sense. When a commercial bank makes a loan, it can create a new deposit at the same time. Banks still face important limits from capital, liquidity, funding costs, credit risk, regulation, borrower demand and monetary policy. Modern money creation therefore does not follow a fixed reserve-multiplier process.

What is the difference between M1 and M2?

M1 generally represents a narrower measure of highly liquid money, while M2 adds some less-liquid assets that remain close substitutes for money. The exact components vary between monetary systems. For example, the Federal Reserve's definitions are not identical to the monetary aggregates used by the European Central Bank.

Is Bitcoin actually money?

Bitcoin can perform some monetary functions, including transferring value between users. However, its limited use as a unit of account, narrower payment acceptance and substantial price volatility mean that it does not perform all the functions of established currencies equally well. Stablecoins are a different case because they usually aim to track an existing monetary unit such as the US dollar.

Does an increase in the money supply automatically weaken a currency?

No. Exchange rates are relative prices between currencies and respond to many factors, including monetary-policy expectations, inflation, interest-rate differentials, capital flows and risk conditions. Markets may also price expected policy changes in advance. A change in one monetary aggregate therefore does not imply a mechanical or proportional change in the exchange rate.

References

International Monetary Fund — Money: At the Center of Transactions

Overview of money as a medium of exchange, unit of account and store of value, including the double-coincidence-of-wants example.

Read at the IMF

Bank of England — Money Creation in the Modern Economy

Explains how commercial-bank lending creates deposits and why modern money creation does not follow the mechanical textbook reserve-multiplier model.

Read at the Bank of England

Federal Reserve — Money Stock Measures (H.6)

Official US definitions and current data for M1, M2, currency, reserves and the monetary base.

View the Federal Reserve H.6 release

European Central Bank — What Are Monetary Aggregates?

ECB methodology for euro-area M1, M2 and M3 and the role of liquidity in determining the boundaries of monetary aggregates.

View ECB methodology

International Monetary Fund — Digital Money, Cross-Border Payments, International Reserves, and the Global Financial Safety Net

IMF framework for distinguishing money from unbacked cryptoassets and stablecoins using the monetary functions of store of value, unit of account and medium of exchange.

Read at the IMF

Bank for International Settlements — The Impact of Stablecoins on the International Monetary and Financial System

2026 analysis of stablecoins as potential stores of value and media of exchange, including their strong concentration in US dollar-denominated instruments.

Read BIS Paper No. 170

International Monetary Fund — The Crypto Cycle and US Monetary Policy

Empirical analysis of the common factor in crypto prices and the relationship between crypto markets, global risk-taking and US monetary-policy tightening.

Read IMF Working Paper 2023/163

Federal Reserve — The Sensitivity of the U.S. Dollar Exchange Rate to Changes in Monetary Policy Expectations

Research on exchange-rate responses to changes in expected US and foreign monetary policy, including the importance of policy surprises and relative rate expectations.

Read at the Federal Reserve

Stylized Facts of Cryptocurrency Markets: Robust Definitions and Inference Approaches

2026 peer-reviewed evidence that Bitcoin and Ethereum returns exhibit heavy tails, little linear return autocorrelation and volatility clustering broadly similar to developed financial markets.

Read in Emerging Markets Review

Lux & Schornstein — Genetic Learning as an Explanation of Stylized Facts of Foreign Exchange Markets

Peer-reviewed treatment of well-established FX time-series features, including unit-root-like behaviour in exchange-rate levels, fat tails in returns and volatility clustering.

Read in the Journal of Mathematical Economics

Sources checked September 1, 2026. Finansified prioritizes central banks, international financial institutions and peer-reviewed research. Important claims are also linked contextually within the article where appropriate.