faqs.org » Abstracts Index » Abstracts » Business, general

Standardized time series L(sub p)-norm variance estimators for simulations

Article Abstract:

A group of standardized time series estimators for the variance parameter sigma(super 2) = lim(sub n->infinity) n Var(Y(sub n)) is presented. These estimators have a very close similarity with L(sub p) norms of Brownian bridges. They generalize several previously analyzed estimators of sigma(super 2), namely, the unweighted area estimator of L.W. Schruben and the unweighted CvM estimator of D. Goldsman et.al. The proposed estimators are all asymptotically unbiased for sigma(super 2) but their finite-sample bias cannot be ignored. On the positive side, the new estimators result into substantial asymptotic variance decreases in contrast to the unweighted area estimator. Hence, the new estimators are more efficient than the previous estimators when the sample size is large enough.

Analysis, Analysis of variance, Time-series analysis, Time series analysis, Stationary processes

---

Consensus forecasts of corporate earnings: analysts' forecasts and time series methods

Article Abstract:

A forecasting method is developed which averages the forecasts made by different methods. Corporate earnings per share are predicted using combinations of financial analyst's forecasts and time series models. The forecasts of analysts are more accurate than time series methods for forecasting horizons of less than one year. Neither method is substantially better than random walk predictions made near the beginning of a fiscal year. The results of the study suggest that there are some benefits to combining analysts' forecasts and time series methods when predictions are made during the first half of the year.

Research, Management science, Forecasting, Financial research

---

Some parts © 2012 Advameg, Inc.