Academic journal article Review - Federal Reserve Bank of St. Louis

Resolving the Liquidity Effect

Academic journal article Review - Federal Reserve Bank of St. Louis

Resolving the Liquidity Effect

Article excerpt

The effect on interest rates of a change in monetary policy has long been an important topic in monetary economics, and there is now a large body of literature that has studied the existence and magnitude of any such effect. Strong conclusions have emerged, and yet, little is available by way of work that attempts to account for the diversity of conclusions. This article aims to fill some of this gap. As the title suggests, it does this by separating out the basic elements of the arguments that lead to the recorded conclusions. In later sections, these are enumerated and discussed. The first section of the article sets out the framework underlying existing studies, followed by an examination of whether the proper object of investigation is a single relationship or a complete system. We come down in favor of the systems viewpoint. Even then, there are many other factors that can account for a diversity of outcomes, and section three is devoted to a consideration of these, ranging from issues of measurement to the sample of data selected for the empirical work. The fourth section explores the interrelationship of monetary policy and the term structure, while the final section presents some conclusions.

THE BASIC MODEL

Although there has been some dissent over the years, mainly from those believing that excess money balances have a powerful direct influence on expenditures, conventional wisdom on the transmission mechanism of monetary policy has been that the effects are felt via interest rates. A very stylized view of this mechanism is available from the money demand and supply relations, which are either explicit or implicit in most models:

(1) (Equation omitted)

(2) (Equation omitted)

where d indicates demand, s supply, m sub t is the log of nominal money, r sub t is the nominal interest rate, while (Equation omitted) and (Equation omitted) are mutually uncorrelated demand and supply shocks. In the textbook treatment of this model, r sub t , responds to shifts in the money supply, engineered by varying beta sub 1 , and the relation dr sub t /dbeta sub 1 = (alpha sub 2 - beta sub 2 ) sub -1 means that the interest rate decreases when money supply increases, provided alpha sub 2 < 0 and beta sub 2 <= - alpha sub 2 . his negative reaction of the interest rate to a rise in money supply is termed the liquidity effect.

When there is a random variable attached to money supply, a change in beta sub 1 can be thought of as a movement in the expected value of (Equation omitted), and the money supply shock might simply be re-labeled (Equation omitted) with the conceptual experiment performed by changing the expected value of (Equation omitted) from beta sub 1 to a new value. Since, mathematically, there is no difference between the response to a change in (Equation omitted) or a change in the expected value of (Equation omitted) we will henceforth concentrate upon describing the effects of a change in (Equation omitted). Such an orientation is now standard in the literature and will be adopted here, so that the liquidity effect will focus upon the simulated response of interest rates to a money supply shock, setting all other shocks to zero.

The above model is static and implies that all adjustments are instantaneous. To make it dynamic, one might augment each relation in equations 1 and 2 with lagged values in m sub t and r sub t to produce

(3) (Equation omitted)

with (Equation omitted) being polynomials in the lag operator of the form (Equation omitted) There is now a distinction to be made between impact effects and the responses over time. In general, one can solve these equations to produce a moving-average representation for interest rates:

(4) (Equation omitted)

where C sub j (L) = (c sub oj + C sub ij L+ ....), and the impact effect will be C sub os = (alpha sub 2 - beta sub 2 ) sup -1 while the effects over time are measured from the impulse responses c sub ks . …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.