Why hold money? The demand for money

In the last module, money was something banks create. Now turn it round: how much of their wealth do people choose to hold as money, rather than as bonds, shares or houses? Money is the only asset you can spend directly, but cash earns nothing and a current account very little. Holding it has a price: the interest you could have earned on a bond instead. That price, the opportunity cost of holding money, is the nominal interest rate (Hyderabad 2013).

Two features of the question matter throughout. It is about a stock, the money you hold at a moment, not a flow of spending. And what people care about is what money buys, so the demand is for real balances, M/P. If every price doubles, people want to hold twice as many rupees and the same real balances.

Irving Fisher (1911) started from an identity. In a year, the rupees spent on transactions must equal the value of the transactions. If the money stock M changes hands V times a year, then

As written, this is true by definition. It becomes a theory once you add two assumptions: velocity is fixed by payment habits and institutions, and output is fixed at full employment by real forces. Then M and P move together: double the money stock and the price level doubles (Hyderabad 2014, 2015). Hyderabad 2013 asked for exactly these two assumptions; CUET 2020 asked what the equation needs to explain short-run prices: that T stays unchanged.

In growth rates, MV = PY becomes %ΔM + %ΔV = %ΔP + %ΔY. That is how a central bank that targets money growth sets its target (Hyderabad 2020): with inflation at 4%, real growth expected at 6% and an income elasticity of money demand of 1, money should grow at 4 + 1 × 6 = 10%.

Marshall and Pigou at Cambridge asked a different question: not how fast money circulates, but how much of their income people want to keep as money. Their answer was a fraction k of nominal income:

Algebraically this is MV = PY again, but the change of view is the start of modern monetary theory: money is now something people demand, and k can depend on things, such as the interest rate. With velocity 3, people hold a third of a year’s income as money (Hyderabad 2015). In the classical model money only affects prices: that is the neutrality of money (JNU 2018), which Lesson 3 of the foundations module showed with the classical dichotomy.

Keynes (1936) called the demand for money liquidity preference and gave three reasons for it:

• Transactions: to bridge the gap between receiving income and spending it. Rises with income.
• Precautionary: to meet unforeseen needs, an illness or a sudden opportunity. Also rises with income.
• Speculative: to avoid losses on bonds when interest rates are expected to rise. Falls as the interest rate rises. This is the new part, and the clean break with the classical theory (UGC NET 2026).

So money demand depends on income and the interest rate (GATE 2026), rising with the first and falling with the second (CUET 2022; JAM 2026). A higher price level raises nominal money demand; with the money supply fixed that pushes the interest rate up (Hyderabad 2021), the mechanism behind the AD curve in Lesson 6.

Keynes also changed what sets the interest rate. For the classics, it balanced saving and investment. For Keynes it is the price that balances the demand for money with its supply, and so also the demand for bonds with their supply, since money and bonds are the only two assets (CUET 2025).

Why should the interest rate make people want to hold less money? Start with a fact about bonds. A perpetuity (Keynes had the British “consol” in mind) pays a fixed ₹100 a year for ever. If the market interest rate is 5%, the bond is worth ₹2,000, because ₹2,000 at 5% also earns ₹100. If the market rate rises to 10%, nobody will pay more than ₹1,000 for it. Bond prices and interest rates move in opposite directions.

Now the speculative motive. Keynes assumed each person has a normal rate in mind, and expects the market rate to return to it: the regressive expectations hypothesis (UGC NET 2024). When today’s rate is low, below the normal rate, people expect it to rise, and so expect bond prices to fall. If the fall would cost more than the interest the bond pays, they hold money instead. The lower the rate, the more people are in that position. That is why speculative demand for money is large when the current rate is lower than people expect it to be (JNU 2021).

At a very low interest rate almost everyone expects the rate to rise, so bond prices can only fall. Nobody wants bonds; everyone would rather hold money. Any extra money the central bank supplies is simply held as idle balances (UGC NET 2025), and the interest rate cannot be pushed lower. The money demand curve turns horizontal: its interest elasticity is infinite (CUET 2020; Hyderabad 2019). Keynes called this the liquidity trap, and in it monetary policy cannot lower the interest rate and so cannot stimulate spending (JAM 2025; Hyderabad 2016).

Keynes made transactions demand depend on income only. Baumol (1952) and Tobin (1956) showed that it depends on the interest rate too. Treat cash like a shop’s inventory. Your year’s spending Y sits in a savings account earning interest i. Each trip to the bank costs b (time, fees) and brings back Y/n. Few trips mean big cash balances and lots of lost interest; many trips mean small balances and lots of trip costs. Find the cheapest number of trips below.

Three exam facts come out of this. The optimal number of trips rises with the interest rate and income and falls with the cost of a trip (CUET 2023). Money demand has an income elasticity of ½ and an interest elasticity of −½ (DSE 2005): there are economies of scale in holding cash. And transactions demand is interest-elastic after all, which is the model’s advantage over Keynes; it does not keep Keynes’s split between transactions and speculative money, which is why “it has a dichotomy between transaction and speculative demand” is the false claim in UGC NET 2026.

Tobin (1958) rebuilt the speculative motive without Keynes’s fixed normal rate. Bonds pay interest but their price is risky; money is safe. A risk-averse person holds a mix, and a higher interest rate tilts the mix towards bonds. This is “liquidity preference as behaviour towards risk”, and it gives each person a smooth demand curve.

Friedman (1956) restated the quantity theory as a theory of the demand for money: money is one asset among many (bonds, equities, physical goods, human wealth) in which wealth is held, and its demand is a stable function of permanent income and the returns on the other assets (JNU 2019). Stability was the point: if money demand is stable, changes in the money supply have predictable effects on nominal income. That is where Friedman challenged the Keynesians (Hyderabad 2019), and it is partly Keynesian (a portfolio choice) and partly not (an emphasis on stability) (UGC NET 2023).

Ordering questions are common, and two in our bank have keys that contradict these dates. CUET 2026 Q46’s key puts Tobin’s portfolio model before Friedman’s restatement, and UGC NET June 2024 Q47’s puts Friedman before Baumol. Neither has an option in the true order, so both are left out of the practice sets. UGC NET December 2024 (Fisher, Cambridge, Baumol, Tobin) is keyed correctly.