A importância dos dados setoriais na previsão financeira

: aplicação ao ROA

Abstract

In this work we investigate the importance of the use of sector data in the financial forecasting of the indicator Return on Assets (ROA) for a sample of 291 firms in the footwear manufacturing sector in Portugal. To measure the impact of sector data in the forecasting process we use five historical data sets: i) of the company, ii) of the sector, iii) combination between company and sector data and iv) of the subsector, to obtain forecasts from a simple forecasting method: the percentage-of-sales. We used this method in order to obtain an approximation to the forecasting methods used in Small and Medium Enterprises (SME) (Armstrong, 2009; Dalrymple, 1987) which represented more than 99,7% of the Portuguese business sector in 2008 (INE, 2010). Our results suggest that the use of sector data can improve the accuracy of ROA forecasts, in particular through the combination of sector and company historical data. The use of sector data, relative to the use of company data, obtained more accurate forecasts for Sales and Assets Turnover and less accurate forecasts for the Gross Margin Rate, Cost Control Effect and Operating Return on Sales. These results suggest that both Sales and Adjusted Assets values converge to the sector while the Gross Margin, Operational Result and the difference between them don‟t converge. Sorting the sample by firm size, using the Adjusted Assets value as a criterion, the obtained results suggest that as firm size diminishes the predictive ability of sector data increases. This work fills a gap in the Corporate Finance literature, particularly in the Financial Forecast subject, with respect to the use of sector data in order to increase forecasting accuracy.

Date of Award	Nov 2011
Original language	Portuguese
Awarding Institution	Universidade Católica Portuguesa
Supervisor	Luis Pacheco (Supervisor)