Lagos–Wright Monetary Model

The Lagos–Wright monetary model is a tractable general-equilibrium framework in business and economics theory for studying money, prices, and monetary policy from first principles. Developed by Ricardo Lagos and Randall Wright in their 2005 Journal of Political Economy article "A Unified Framework for Monetary Theory and Policy Analysis", it merges the search-theoretic approach of the Kiyotaki–Wright model with the competitive markets of general equilibrium theory. Earlier search models forced agents to hold only one indivisible unit of money, which made prices and inflation hard to analyze. The Lagos–Wright design lets money be divisible and agents hold arbitrary balances, yet keeps the distribution of balances simple enough to solve in closed form.

Time is discrete and infinite. Every period has two sub-periods. In the first, a decentralized market (DM), agents meet in random bilateral matches. Buyers wish to consume a good the seller can produce, but trade is subject to frictions: partners are anonymous, there is no record-keeping, and no commitment to repay. A medium of exchange is therefore necessary, since credit is not feasible between strangers. This is the same type of friction highlighted by information asymmetry and limited enforcement in contract theory.

In the second sub-period, a centralized market (CM), all agents trade together in a competitive Walrasian market. They consume a general good X, supply labor H to produce it, and rebalance their money holdings at the price of money φ_t in terms of the general good.

Agents have preferences over three objects: DM consumption q, CM consumption x, and labor h. A typical specification is

U(q, x, h) = u(q) + x − h,

which is quasi-linear in the CM good and labor. Money is an intrinsically worthless, perfectly divisible, storable asset supplied by a central bank that grows its stock at gross rate μ via lump-sum transfers.

The central technical insight concerns the CM portfolio decision. Let W(m) denote the value of entering the CM with m units of money. Because utility is linear in x and h, the CM value function is linear in wealth: W(m) = φm + W(0). Consequently, the amount of money an agent chooses to carry out of the CM is independent of the money she carried in. Every agent, whatever her trading history in the DM, leaves the CM with the same balance. The cross-sectional distribution of money holdings at the start of the next DM is therefore degenerate: all agents hold the same amount.

This result removes the need to track a complicated distribution as a state variable, which had made previous divisible-money models intractable. Heterogeneity arises only transiently inside the DM, through random matching, and is wiped out when agents rebalance. The technique is a major reason the framework became the workhorse of New Monetarist economics.

In a DM match, a buyer with money balance m and a seller who can produce at cost c(q) determine the quantity q and payment d ≤ m. Lagos and Wright employ Kalai bargaining, a generalization of Nash bargaining in which the buyer receives share θ of the match surplus subject to the feasibility constraint that she cannot pay more than she holds. Other protocols are used in the literature, including take-it-or-leave-it offers (θ = 1), Walrasian price taking in competitive search, and proportional bargaining.

When the cash constraint binds, the buyer spends all her balances: d = m. The DM quantity then rises with money holdings, q = q(m), with a slope that depends on θ and the curvature of u. The marginal value of a unit of money in the DM is a liquidity premium: money is valued for what it can buy in the DM, over and above its role as a store of value.

The first-order condition for the optimal choice of next-period balances in the CM yields an Euler equation of the form

φ_t = β φ_t+1 [ 1 + λ ( u′(q_t+1) / c′(q_t+1) − 1 ) ],

where β is the discount factor, λ is the DM trading probability, and the bracketed term includes the liquidity wedge from the possibility of a DM trade. In a stationary monetary equilibrium, φ_t+1 / φ_t = 1 / μ. The equilibrium determines DM output q as a decreasing function of the money growth rate μ: higher inflation acts as a tax on carrying money, so agents bring fewer balances and trade less. The model also generates a non-monetary equilibrium (q = 0) alongside the monetary one, reflecting the self-fulfilling nature of money seen in the Nash equilibrium of coordination games.

Milton Friedman argued that the optimal policy sets the nominal interest rate to zero, so that the private opportunity cost of holding money equals its near-zero social cost of production. In the Lagos–Wright model this condition emerges endogenously. When the nominal interest rate is zero (μ = β), the liquidity wedge vanishes and DM trades reach the efficient quantity if buyers hold the surplus-maximizing share (θ = 1). When θ < 1, the efficiency requires the buyer to capture the full surplus, a standard holdup problem related to moral hazard in bilateral relationships. In such cases even the Friedman rule is insufficient for a first-best outcome, and Pareto efficiency can only be approached through a negative inflation rate or alternative trading mechanisms.

The model allows quantification of the welfare cost of inflation: the fraction of consumption agents would give up to move from a given inflation rate to the Friedman rule. Calibrations to US data by Lagos and Wright suggest costs of several percent of consumption at moderate inflation rates, which is larger than the classic estimates based on money-demand area calculations by Robert Lucas (2000). The difference arises because bargaining and search externalities amplify the distortion, a point developed in Lagos, Rocheteau, and Wright (2017).

The tractability of the framework has supported many extensions:

• Credit and banking: adding limited commitment to study how credit competes with and complements money.
• Asset pricing and liquidity: allowing assets other than money, including government bonds, equity, and housing, to serve as collateral, with prices reflecting liquidity premia.
• Monetary policy: analysis of open-market operations, interest on reserves, and quantitative easing as shifts in the supply of liquid assets.
• Labor and unemployment: embedding Mortensen–Pissarides search frictions alongside monetary frictions.
• Mechanism design: characterizing which trading arrangements are feasible when money is the only record of past trades, which links the model to mechanism design.

The degenerate distribution is a consequence of quasi-linearity, which eliminates wealth effects in the CM. Relaxing it brings back distributional dynamics that must be solved numerically. The alternating-market structure is also stylized, and the model's predictions about the velocity and cyclical behavior of money depend on the calibrated DM matching function. Nonetheless, the framework remains the standard microfoundation for studying why people hold money and how inflation erodes the value of exchange.

1. Lagos, R., & Wright, R. (2005). A Unified Framework for Monetary Theory and Policy Analysis. Journal of Political Economy, 113(3), 463–484. https://doi.org/10.1086/429804
2. Kiyotaki, N., & Wright, R. (1989). On Money as a Medium of Exchange. Journal of Political Economy, 97(4), 927–954. https://doi.org/10.1086/261634
3. Lucas, R. E. (2000). Inflation and Welfare. Econometrica, 68(2), 247–274. https://doi.org/10.1111/1468-0262.00109
4. Lagos, R., Rocheteau, G., & Wright, R. (2017). Liquidity: A New Monetarist Perspective. Journal of Economic Literature, 55(2), 371–440. https://doi.org/10.1257/jel.20141195