Larry W. Taylor

Diagnostics for IV Regressions

The use of residuals from the structural equations in a simultaneous-equations model can lead to misleading measures of association and to invalid diagnostic statistics for heteroscedasticity and functional form misspecifications. The issue of appropriate measures of goodness-of-fit for IV regression is addressed in Pesaran and Smith (1994). This paper considers the problem of appropriate construction of diagnostics for IV regressions, and proposes simple tests for functional form misspecifications and heteroscedasticity.

Swap Covered Interest Parity in Long-Date Capital Markets

Using the currency swap as the forward-exchange risk hedge, the covered interest parity condition in the long-date capital markets is evaluated. Of interest is the extent to which deviations from parity can be attributed to transactions costs. The empirical conclusions presented in the paper suggest that, although (on average) transactions costs account for deviations from parity, net deviations (in excess of transactions costs) are neither rare nor short-lived. Yet an analysis of the variance structure of covered interest parity reveals that these profit opportunities diminish over time and eventually disappear. Copyright 1996 by MIT Press.

The size bias of White's information matrix test

Abstract A recently proposed diagnostic test for maximum-likelihood estimators is examined. It is found that the size bias is very substantial and this makes the test impractical to use in its current form.

Electricity demand in multi-family, renter-occupied residences

This paper focuses on multi-family, renter-occupied housing where the landlord provides major appliances, a large residential sector that has not received widespread attention in the literature. Experimental data are used from a test where one group of households was billed directly for the electricity they consumed, while a control group had unlimited electric service included as part of the monthly rent. The empirical analysis yields short-run demand estimates that satisfy to a remarkable extent the ceteris paribus assumptions of demand analysis. This is the case for two reasons: (1) the households did not individually execute the appliance choice decision at various (unknown) times in the past, and (2) the stock of appliances upon which the derived short-run demand is conditional is identical for all households in the sample. A second important feature is emphasis on functional form. Examined in some detail is the question of the appropriate functional form for residential electricity demand. Finally, empirical results of the analysis are used to investigate the cost-effectiveness and welfare effects of installing submeters in existing multi-family residences.

A non-parametric analysis of covered interest parity in long-date capital markets

Abstract The financially innovative currency swap has elevated the international mobility of long-term assets. Of interest is whether the covered interest parity condition now holds in these markets, with the currency swap as the forward-exchange risk hedge. The empirical conclusions presented in the paper suggest that deviations from covered interest parity (in excess of transactions costs) are not rare. Yet, while profit opportunities are not always short-lived, an analysis of the deviations in excess of transactions costs reveals that they diminish over time, and eventually disappear. (JEL F32).

Student Personality Type Versus Grading Procedures in Intermediate Accounting Courses

Abstract Though teams composed of various personality types are expected to perform quite well as problem solvers, the extant literature has demonstrated a fairly narrow range of personality types in business courses and the accounting profession. To determine whether various grade-influencing activities in accounting courses favor certain personality types, the personality preferences and temperaments of 82 intermediate accounting students were determined. Relationships were then found between personality variables and the number of class absences, class participation, and the performance on homework and problems on the final examination.

A Nonparametric Investigation of Duration Dependence in the American Business Cycle: A Note

A recent empirical investigation by Diebold and Rudebusch (1990b) into the nature of business cycles reveals an apparent contradiction concerning the duration dependence of half and whole cycles. While expansions and contractions appear to be duration independent, completed cycles seem to be duration dependent. This implies, for example, that a long expansion is no more likely to terminate than a short expansion, whereas the termination probability for a whole cycle depends on its length. As discussed in Diebold and Rudebusch, this perplexing situation is feasible if the lengths of expansions and contractions are negatively correlated. However, they find only weak empirical evidence to support this conjecture. The purpose of this note is to resolve the puzzle by recognizing that it is impossible for expansions, contractions, and completed cycles to all follow the constant-hazard distribution. We then conclude that the (apparent) contradiction arises because the same statistical tests are applied to both half and whole cycles. To illustrate, because data are reported on a monthly basis, we formulate two discrete-time tests for duration dependence. The first of these is applied separately to expansions, contractions, and completed cycles; the second is employed to test jointly the null hypothesis that expansions and contractions are both duration independent. To derive the tests, initially assume that expansions and contractions have constant transition probabilities. If X is the random variable for the number of consecutive months of (say) expansion, p is the (constant) transition probability, and te represents the minimum

A Benchmark Profile of Economics Departments in 15 Private Universities

Abstract During the spring of 1999, the authors completed a benchmarking survey of 15 economics departments in private universities as part of a strategic planning exercise. All are selective medium-sized institutions that experience roughly the same types of market pressures and compete for the same types of students. The authors report the information gleaned from the survey concerning such items as departmental resources, teaching loads, class sizes, departmental research expectations, and weights given to research, teaching, and service in salary determination and promotion. The authors believe that their results and methods might be useful to other economics departments engaging in bench-marking exercises.

Some non-parametric tests for duration dependence: an application to UK business cycle data

A distinction between Fishers implied data-generating process for Monte Carlo cycles and the more general Markov process leads to non-parametric tests for duration dependence. Tests are based on the method of moments, Tauchens generalized method of moments (GMM) procedure, and a statistic whose null distribution probability limit is zero. Using finite-sample critical values obtained by Monte Carlo methods, our test results are remarkably consistent. The null distribution of the GMM test statistic for samples of the size considered is distinctly non-normal, so that asymptotic critical values give erroneous results. The tests are applied to UK business cycle data for 1854–1992. There is evidence for duration dependence in expansions but not in contractions.

An R2criterion based on optimal predictors

The predictor that minimizes mean-squared prediction error is used to derive a goodness-of-fit measure that offers an asymptotically valid model selection criterion for a wide variety of regression models. In particular, a new goodness-of-fit criterion (cr2) is proposed for censored or otherwise limited dependent variables. The new goodness-of-fit measure is then applied to the analysis of duration.