𝔖 Bobbio Scriptorium
✦   LIBER   ✦

Quarterly GPPA earnings data: Time-series properties and predictive ability results in the airlines industry

✍ Scribed by W. A. Hillison; W. S. Hopwood; K. S. Lorek


Publisher
John Wiley and Sons
Year
1983
Tongue
English
Weight
830 KB
Volume
2
Category
Article
ISSN
0277-6693

No coin nor oath required. For personal study only.

✦ Synopsis


This paper provides preliminary results regarding the impact of general purchasing power adjustments (GPPA) on quarterly earnings for 24 firms from the airlines industry. Findings indicated that GPPA transformations do not substantially alter the time-series properties of quarterly earnings data. Two Box-Jenkins models, (100) x (100) and (100) x (110) were identified as possible parsimonious models for the airlines industry. It I s recommended that these structures be considered as viable candidates for earnings expectations models in market studies testing for informational content of GPPA earnings. The predictive findings demonstrated that predictions of historical cost quarterly earnings were significantly more accurate than GPPA predictions for four of the five horizons tested.

KEY WORDS Box-Jenkins models Reported quarterly earnings

Purchasing power adjusted earnings Predictive ability

This paper provides preliminary descriptive and predictive evidence on the time-series properties of general purchasing power adjusted (GPPA) data. The impact of these adjustments on the quarterly income series is assessed by comparing the time-series models for GPPA and historical cost (HC) earnings using goodness of fit criteria. Predictive ability tests of GPPA and HC earnings data are also provided. Empirical evidence on the descriptive and predictive ability of the GPPA models should be of interest primarily to researchers in accounting and finance who wish to select expectations models for GPPA earnings.

BACKGROUND

Research on the time-series properties of alternative income measures relates to several study areas in accounting. These include (1) informational content and association, (2) income-smoothing, (3) security valuation and (4) forecasting. The contribution of the present study is assessed after discussing briefly the linkages between statistical time-series research and each of the aforementioned areas.