Detecting shocks: Outliers and breaks in time series

A single outlier in a regression model can be detected by the effect of its deletion on the residual sum of squares. An equivalent procedure is the simple intervention in which an extra parameter is added for the mean of the observation in question. Similarly, for unobserved components or structural time-series models, the effect of elaborations of the model on inferences can be investigated by the use of interventions involving a single parameter, such as trend or level changes. Because such time-series models contain more than one variance, the effect of the intervention is measured by the change in individual variances.We examine the effect on the estimated parameters of moving various kinds of intervention along the series. The horrendous computational problems involved are overcome by the use of score statistics combined with recent developments in filtering and smoothing. Interpretation of the resulting time-series plots of diagnostics is aided by simulation envelopes.Our procedures, illustrated with four example, permit keen insights into the fragility of inferences to specific shocks, such as outliers and level breaks. Although the emphasis is mostly on parameter estimation, forecast are also considered. Possible extensions include seasonal adjustment and detrending of series.     

On compensation for risk aversion and skewness affection in wages

This paper presents extensive empirical testing of the hypothesis that greater post-schooling earnings risk requires higher expected returns. Expanding on this notion, on the basis of utility theory, we predict that workers not only care about risk but also about the skewness in the distribution of the compensation paid: workers exhibit risk aversion and skewness affection. To test these hypotheses, this paper carefully develops various measures of risk and skewness by occupational/educational classification of the worker and finds supportive evidence: for men, wages rise with occupational earnings variance and decrease with skewness, for women only the negative effect of skewness is significant.     

If You Are So Smart,Why Aren't You an Entrepreneur? Returns to Cognitive and Social Ability: Entrepreneurs Versus Employees

How valuable are cognitive and social abilities for entrepreneurs’ relative to employees’ earnings? We answer three questions: (1) To what extent does a composite measure of ability affect an entrepreneur's earnings relative to wages earned by employees? (2) Do different cognitive abilities (e.g., math ability, language, or verbal ability) and social ability affect earnings of entrepreneurs and employees differently?, and (3) Does the balance in these measured ability levels affect an individual's earnings? Our (difference‐of‐difference) estimates of the returns to ability for spells in entrepreneurship versus wage employment account for selectivity into entrepreneurial positions insofar as they are determined by fixed individual characteristics. Our robust results provide the following answers to the three questions: General ability has a stronger impact on entrepreneurial incomes than on wages. Moreover, entrepreneurs and employees benefit from different sets of specific abilities: verbal and clerical abilities have a stronger impact on wages, whereas mathematical, social, and technical ability are more valuable for entrepreneurs. The balance in the various kinds of ability also generates a higher income, but only for entrepreneurs: This finding supports Lazear's Jack‐of‐all‐Trades theory.     

Periodic Unobserved Cycles in Seasonal Time Series with an Application to US Unemployment*

We introduce a general class of periodic unobserved component (UC) time series models with stochastic trend and seasonal components and with a novel periodic stochastic cycle component. The general state space formulation of the periodic model allows for exact maximum likelihood estimation, signal extraction and forecasting. The consequences for model‐based seasonal adjustment are discussed. The new periodic model is applied to postwar monthly US unemployment series from which we identify a significant periodic stochastic cycle. A detailed periodic analysis is presented including a comparison between the performances of periodic and non‐periodic UC models.     

Estimating domestic content in exports when processing trade is pervasive

For many questions, it is crucial to know the extent of domestic value added (DVA) in a country's exports, but the computation is more complicated when processing trade is pervasive. We propose a method for computing domestic and foreign contents that allows for processing trade. By applying our framework to Chinese data, we estimate that the share of domestic content in its manufactured exports was about 50% before China's WTO membership, and has risen to nearly 60% since then. There are also interesting variations across sectors. Those sectors that are likely labeled as relatively sophisticated such as electronic devices have particularly low domestic content (about 30% or less).     

Evaluating the performance of global emerging markets equity exchange-traded funds

We examine the performance of passively managed exchange-traded funds (ETFs) that provide exposure to global emerging markets equities. We find that the tracking errors of these funds are substantially higher than previously reported levels for developed markets ETFs. ETFs that use statistical index replication techniques turn out to be especially prone to high tracking errors, and particularly so during periods of high cross-sectional dispersion in stock returns. At the same time, we find no convincing evidence that these funds earn higher returns than ETFs that rely on full-replication techniques.     

Compensation for Earnings Risk under Worker Heterogeneity

We use two large data sets to estimate the Risk Augmented Mincer equation and test for risk compensation in expected earnings. We replicate earlier findings of a positive premium for risk and a negative premium for skew and add confirmation of the key results if we control for individual ability. We compare risk compensation and risk distributions between some labor market groups and find that immigrants and natives do not differ in risk attitudes, that public sector workers are undercompensated for their risk, and that risk compensation by gender is not fully consistent with higher risk aversion for women. We express concern that a linear compensation model may be too simple.     

Earnings functions: Testing for the demand side

To analyze labor market phenomena, one may simultaneously grade workers (by education) and jobs (by level of difficulty and complexity). Human capital theory predicts earnings based on only worker quality. Thurow's version of a segmented labor market theory predicts that only the job level is relevant, and the assignment theory predicts that both variables are relevant. Using three different data sets for The Netherlands, testing demonstrates convincingly the superiority of the assignment theory.     

Cross-sectional learning and short-run persistence in mutual fund performance

Using monthly return data of more than 6400 US equity mutual funds we investigate short-run performance persistence over the period 1984–2003. We sort funds into rank portfolios based on past performance, and evaluate the portfolios’ out-of-sample performance. To cope with short ranking periods, we employ an empirical Bayes approach to measure past performance more efficiently. Our main finding is that when funds are sorted into decile portfolios based on 12-month ranking periods, the top decile of funds earns a statistically significant, abnormal return of 0.26 percent per month. This effect persists beyond load fees, and is mainly concentrated in relatively young, small cap/growth funds.