Earnings Forecasts and Stock Price Data: How stock prices can be used to forecast less biased and more accurate earnings

Sammanfattning: In this study, we investigate the effect of incorporating stock price data in model-based forecasts. Previous research shows that stock price data can be used to forecast less bias and more accurate earnings. We expand the literature of earnings forecasts on Swedish data and our results are suggestive that stock price can be used to forecast less biased, but not more accurate, earnings. Our findings also indicate that the stock price contains incremental information about future earnings in the short-term, compared to pure-accounting based models and forecasts by financial analysts. Furthermore, we also look into the effect that firm characteristics have on forecast performance. The results are suggestive that the forecast accuracy for model-based forecasts is dependent on firm characteristics, such as industry membership, size, and earnings-to-price ratio.