آزمون ارزش فعلی عادت ها و حساب جاری

Forecasts derived from standard intertemporal current account (ICA) models generally fail to match the volatility of actual current accounts. This paper offers a solution to the “excess volatility” problem of standard ICA models by incorporating consumption habits into the standard model. The model, as developed in the paper, shows that significant habit formation implies increased current account volatility, as sluggishness is introduced into the consumption adjustment process that follows income shocks. A theory-consistent measure of the degree of habit formation is estimated using GMM. The estimated habit parameter is found to be statistically significant in six of eight quarterly samples.

A central feature of intertemporal current account (ICA) models, as introduced by Sachs (1981), is the important role attributed to aggregate consumption behavior in the determination of the current account. An implication of this is that the current account predictions of any particular ICA model are highly contingent on the model's specification of aggregate consumption behavior. In the standard ICA model, consumption is modeled as being chosen by an infinitely-lived representative agent with a time separable period utility function. This paper examines the model implications of relaxing the assumption of time separable utility through preferences that allow for habit formation in consumption.1 Habits are shown to significantly improve the empirical performance of the ICA model. Particularly, current account forecasts derived from the habit model better match the volatility of actual current accounts.2 The paper also provides open-economy estimates of the degree of habit formation in aggregate consumption for eight countries. Habit formation affects the current account response to an income shock by slowing the adjustment of consumption to the shock. In the standard model, assuming a rate of time preference equal to the interest rate, consumption is set equal to permanent income and the current account is completely determined by deviations between current income and permanent income. With habit formation, the slow adjustment of consumption to a shock creates a temporary gap between consumption and permanent income that affects the current account in addition to any difference between current income and permanent income. Income shocks that affect current income to an equal or greater degree than permanent income have exaggerated current account effects with habits, and thereby increase the current account volatility that results from such shocks. The degree of habit formation is estimated through the imposition of model-implied restrictions on the parameters of a VAR comprised of the current account and income changes, in a modification of the present value test methodology of Campbell (1987).3 The VAR provides an estimate of the data-generating process for income that then allows the degree of habit formation consistent with a country's observed current account to be estimated. The remainder of this paper takes the following form. The proceeding section presents an intertemporal model of the current account that deviates from the intertemporal separability assumption standard to the literature and allows for habit formation. The implications of habit formation in regard to the dynamics of the current account are then examined. The third section outlines the estimation strategy and results. The fourth and final section concludes.

The estimation results suggest that habits play an important role in determining the dynamics of the current account. The addition of habit formation to the standard model improves the empirical fit of the model in all but one circumstance. The habit-formation model is able to better approximate the volatility of the current account, offering a potential solution to one of the standard model's most consistent failings. By imposing the theoretical restrictions of the habit model onto the data, estimates of the degree of habit formation necessary to match the data are obtained. In all but one sample, the null that the estimates satisfy the orthogonality conditions implied by the model cannot be rejected at standard levels of significance. It is interesting to note the proximity of many of the estimated habit parameters to previous estimates in literature. Habit formation models by Deaton (1987) and Constantinides (1990) require habit parameters of around 0.80 in order to explain aggregate consumption smoothness and the equity-premium puzzle, respectively. For the current study the average estimate of γγ is quite close to 0.80.