Research Database

This paper explains why the risky notes of banks established during the Free Banking Era (1837–63) were demanded even when relatively safe specie (gold and silver coin) was an alternative. Free bank notes were demanded because they were priced to reflect the expected value of their backing. The empirical evidence supports this explanation. Specifically, in New York, Wisconsin, and Indiana the expected value of backing was sufficient for free bank notes to circulate at par, which they did. In Minnesota the backing for notes was very poor: they exchanged well below par, being treated as small-denomination securities.

An examination of the behavior of stock returns around quarterly earnings announcement dates finds a seasonal pattern: small firms show large positive abnormal returns and a sizable increase in the variability of returns around these dates. Only part of the large abnormal returns can be accounted for by the fact that firms with good news tend to announce early. Large firms show no abnormal returns around announcement dates and a much smaller increase in variability.

The consequences of a straightforward monetary targeting scheme are examined for a simple dynamic macro model. The notion of “targeting” used is the strategic one introduced by Rogoff (1985). Numerical calculations are used to demonstrate that for the model under consideration, monetary targeting is likely to lead to a deterioration of policy performance. These examples cast doubt upon the general efficacy of simple targeting schemes in dynamic rational expectations models.

Deaton (1986) has noted that if income is a first-order autoregressive process in first differences, then a simple version of Friedman’s permanent income hypothesis (SPIH) implies that measured U.S. consumption is insufficiently sensitive to innovations in income. This paper argues that this implication of the SPIH is a consequence of the fact that it ignores the role of the substitution effect in the consumption decision. Using a parametric version of the standard model of economic growth, the paper shows that very small movements in interest rates are sufficient to induce an empirically plausible amount of consumption smoothing. Since an overall evaluation of the model’s explanation for the observed smoothness of consumption requires examining its implications for other aspects of the data, the paper also explores some of these.

The statistical significance of variance decompositions and impulse response functions for unrestricted vector autoregressions is questionable. Most previous studies are suspect because they have not provided confidence intervals for variance decompositions and impulse response functions. Here two methods of computing such intervals are developed, one using a normal approximation, the other using bootstrapped resampling. An example from Sims’ work illustrates the importance of computing these confidence intervals. In the example, the 95 percent confidence intervals for variance decompositions span up to 66 percentage points at that usual forecasting horizon.

Doan, Litterman, and Sims (DLS) have suggested using conditional forecasts to do policy analysis with Bayesian vector autoregression (BVAR) models. Their method seems to violate the Lucas critique, which implies that coefficients of a BVAR model will change when there is a change in policy rules. In this paper we construct a BVAR macro model and attempt to determine whether the Lucas critique is important quantitatively. We find evidence following two candidate policy rule changes of significant coefficient instability and of a deterioration in the performance of the DLS method.

A bivariate Granger-causality test on money and output finds statistically significant causality when data are measured in log levels, but not when they are measured in first differences of the logs. Which of these results is right? The answer to that question matters because a finding of no Granger-causality from money to output would substantially embarrass existing business cycle models in which money plays an important role [Eichenbaum and Singleton (1986)]. Monte Carlo simulation experiments indicate that, most probably, the first difference results reflect lack of power, whereas the level results reflect Granger-causality that is actually in the data.

This paper presents a simple counterexample to the belief that policy cooperation among benevolent governments is desirable. It also explains circumstances under which such counterexamples are possible and relates them to the literature on time inconsistency.

The neoclassical growth model studied in Kydland and Prescott [1982] is modified to permit the capital utilization rate to vary. The effect of this modification is to increase the amplitude of the aggregate fluctuations predicted by theory as the equilibrium response to technological shocks. If following Solow [1957], the changes in output not accounted for by changes in the labor and tangible capital inputs are interpreted as being the technology shocks, the statistical properties of the fluctuations in the post-war United States economy are close in magintude and nature to those predicted by theory.

This paper presents a completely worked example applying the frequency domain estimation strategy proposed by Hansen and Sargent [1980, 1981a]. A bivariate, high order continuous time autoregressive moving average model is estimated subject to the restrictions implied by the rational expectations model of the term structure of interest rates. The estimation strategy takes into account the fact that one of the data series are point-in-time observations, while the other are time averaged. Alternative strategies are considered for taking into account nonstationarity in the data. Computing times reported in the paper demonstrate that estimation using the techniques of Hansen and Sargent is inexpensive.

This paper examines the quantitative importance of temporal aggregation bias in distorting parameter estimates and hypothesis tests. Our strategy is to consider two empirical examples in which temporal aggregation bias has the potential to account for results which are widely viewed as being anomalous from the perspective of particular economic models. Our first example investigates the possibility that temporal aggregation bias can lead to spurious Granger causality relationships. The quantitative importance of this possibility is examined in the context of Granger causal relations between the growth rates of money and various measures of aggregate output. Our second example investigates the possibility that temporal aggregation bias can account for the slow speeds of adjustment typically obtained with stock adjustment models. The quantitative importance of this possibility is examined in the context of a particular class of continuous and discrete time equilibrium models of inventories and sales. The different models are compared on the basis of the behavioral implications of the estimated values of the structural parameters which we obtain and their overall statistical performance. The empirical results from both examples provide support for the view that temporal aggregation bias can be quantitatively important in the sense of significantly distorting inference.

We describe a simple environment in which assets of varying qualities may be used for transactions and consumption. The quality of an asset is known to the seller but not the buyer. We show that this feature can generate a negative relationship between the transactions velocities of assets and their rates of return. We also discuss several versions of Gresham's Law which hold in this environment.

We consider the existence of deterministically cycling steady state equilibria in a class of stationary overlapping generations models with sufficiently long (but, finite) lived agents. Preferences are of the discounted sum of utilities type with a fixed discount rate. Utility functions with large coefficients of relative risk aversion which generate strong income effects (relative to substitution effects) and backward bending offer curves are permitted. Lifetime endowment patterns are quite arbitrary. We show that if agents have a positive discount rate, then as agents1 lifespans get large, short period non-monetary cycles will disappear. Further, constant monetary steady states do not exist and therefore, neither do stationary monetary cycles of any period. We then consider the case where agents have a negative discount rate and show that there are robust examples in which constant monetary steady states as well as stationary monetary cycles (with undiminished amplitude) can occur no matter how long agents live.

A two country overlapping generations model is constructed, in which financial intermediation arises endogenously as an incentive compatible means of economizing on monitoring costs. Because of international credit markets. The model is used to generate the existence of transaction costs, money markets in the two countries are segmented and investors have differential access to predictions concerning the role of international intermediation in economic development, and to examine the nature of business cycle phenomena across alternative exchange rate regimes. Disturbances are propagated by a credit allocation mechanism, which also lends a novel flavor to the model's long run properties.

The fiscal year and the calendar year coincide for a large fraction of firms traded in the New York and American Stock Exchanges. It is therefore possible that part of the large positive abnormal return earned by stocks as a group during the first week of trading in January may be due to temporal resolution of uncertainty accompanying the end of the fiscal year. We study this hypothesis by examining whether stocks of firms with fiscal years ending in months other than December also realize positive abnormal returns, following the end of their fiscal years. We find that there are no excess returns for such firms in the first five trading days following the end of the fiscal year.

This paper investigates two methods of approximating the optimal decision rules of a stochastic, representative agent model which exhibits growth in steady state and cannot be expressed in linear–quadratic form. Both methods are modifications on the linear quadratic approximation technique proposed by Kydland and Prescott. It is shown that one of the solution methods leads to bizarre dynamic behavior, even with shocks of empirically reasonable magnitude. The other solution technique does not exhibit such bizarre behavior.

The Diamond-Dybvig model of banking (Journal of Political Economy, 1983) is amended by introducing communication barriers - these being implicit in their model and in most explanations of why people hold so-called liquid assets. These barriers imply the sequential-service constraint that Diamond and Dybvig imposed on private intermediation and have other implications: infeasibility of the policy that Diamond and Dybvig identify with deposit insurance and desirability of dependence of the realized return on deposits on the random order of withdrawals.

In this paper we analyze the constraints imposed by dynamic consistency in a model of optimal taxation. We assume that only distorting taxes are available to finance government consumption. Optimal fiscal policy requires the use of debt to smooth distortions over time. Dynamic consistency requires that governments at each point in time not have an incentive to default on the inherited debt. We consider policy functions which map the history of the economy including the actions of past governments into current decisions. A sustainable plan is a sequence of history-contingent policies which are optimal at each date given that future policies will be selected according to the plan. We show that if agents discount the future sufficiently little and if government consumption fluctuates then optimal sustainable plans yield policies and allocations which are identical to those under full commitment. We contrast our notion of dynamic consistency with other definitions.

We prove the existence of a competitive equilibrium in an overlapping generations model in which each generation has a preference ordering over its own and its descendents’ consumptions. The model is one of pure exchange with many goods in each period and two period lived generations. The bequest from one generation to the next is required to be non-negative and is endogenous. In equilibrium, some sequences of agents of successive generations may be continually “linked” by positive bequests and act as infinitely lived agents. Other sequences of agents may not be so linked and therefore behave as sequences of finite lived agents. We give three examples which illustrate the following points: (i) multiple equilibria may exist some of which resemble those of standard overlapping generations models, whereas in others some sequences of agents behave as if infinitely lived, (ii) multiple steady states of the above two types may exist in which the latter are unstable and the former are stable, and (iii) if agents have preferences given by discounted sums of utilities with different discount rates, then not all sequences of generations can be continually linked and hence behave as infinitely lived agents.

The existence of fixed points for monotone maps on spaces of measures is established. The case of monotone Markov processes is analyzed and a uniqueness and global stability condition is developed. A comparative statics result is presented and the problem of approximation to the invariant distribution is discussed. The conditions of the theorems are verified for the cases of Optimal Stochastic Growth and Industry Equilibrium.

This paper derives a variance bounds test for a broad class of linear rational expectations models. According to this test if observed data accords with the model, then a weighted sum of autocovariances of the covariance-stationary components of the endogenous state variables should be nonnegative. The new test reinterprets its forefather - West's [1986] variance bounds test - and extends its applicability by not requiring exogenous state variables in order to be tested. The possibility of the test's application to nonlinear models is also discussed.

This paper investigates whether there are simple versions of the permanent income hypothesis which are consistent with the aggregate U.S. consumption and output data. Our analysis is conducted within the confines of a simple dynamic general equilibrium model of aggregate real output, investment, hours of work and consumption. We study the quantitative importance of two perturbations to the version of our model which predicts that observed consumption follows a random walk: (i) changing the production technology specification which rationalizes the random walk result, and (ii) replacing the assumption that agents' decision intervals coincide with the data sampling interval with the assumption that agents make decisions on a continuous time basis. We find substantially less evidence against the continuous time models than against their discrete time counterparts. In fact neither of the two continuous time models can be rejected at conventional significance levels. The continuous time models outperform their discrete time counterparts primarily because they explicitly account for the fact that the data used to test the models are time averaged measures of the underlying unobserved point-in-time variables. The net result is that they are better able to accommodate the degree of serial correlation present in the first difference of observed per capita U.S. consumption.

This paper examines the optimal debt contract between lenders and a sovereign borrower when the borrower is free to repudiate the debt and when his decision to invest or consume borrowed funds is unobservable. We show that recurrent debt crises are a necessary part of the incentive structure which supports the optimal pattern of lending.

We construct a sequence of pure exchange, stationary OLG economies in which generations have longer and longer life spans and all agents maximize a discounted sum of utilities with a fixed, positive, and common discount rate. Period utility functions and endowment patterns are subject to mild restrictions and within generation heterogeneity is permitted. We show that: (i) Every sequence of equilibrium interest rates converges to the discount rate. (ii) Eventually every nonmonetary steady state is optimal and a monetary steady state will never exist. (iii) For any agent consumption at any fixed age converges to permanent income evaluated using the utility discount rate.

One purpose of this article is to exposit the relationship between life cycle models (with constructive immortality) and infinitely lived agents models. We use this to point out problems in interpreting data, especially with regard to the use of interest rates in the class of representative agent models when growth in population and per capita variables is taken into account. We also point out some common misconceptions regarding the "volume of trade" in representative agent models and show how to reconcile the savings profile of the representative agent with the life cycle savings profile in a life cycle model.

This paper investigates the role of aggregate shocks on household consumption and labor supply. It posits, estimates and tests a model where the equilibrium behavior of agents sometimes leads them to locate on the boundary of their respective choices sets. The framework is rich enough to nest much previous empirical work on life cycle labor supply and consumption based asset pricing. It also yields a structural interpretation of wage regressions on unemployment. An important feature of our model is that markets are complete. Consequently, aggregate shocks only enter through two price sequences, namely real wages, and a sequence comprising weighted prices for future contingent consumption claims which are ultimately realized. We examine the properties of this latter sequence, whose elements may be represented as mappings from real wages and aggregate dividends.

Our empirical findings may be grouped into three. First, aggregate shocks play a significant role in determining the choices people make. Second, we reject for males some of the restrictions implicit in structural interpretations of wage unemployment regressions. Moreover when these restrictions are imposed, we find wages are countercyclical, but cannot reject the null hypothesis of no effect. Third, the null hypothesis that markets are complete is not invariably rejected. However, the orthogonality conditions associated with the asset pricing equation are rejected, even though our specification of preferences incorporates types of heterogeneity which violate the necessary conditions for aggregating to a representative agent formulation. Finally, we reject the cross-equation restrictions between the labor supply of spouses implied by equilibrium behavior.