| Private investors claim that they are disadvantaged due to the lack of relevant investment risk information in Hong Kong. This study investigates whether accounting-determined risk (AR) measures can be used to predict the perceived risk class of Hong Kong companies. Risk perceptions on 25 profitable large to medium-sized companies were obtained from financial analysts in Hong Kong. The companies were then classified into three classes--low, medium and high--according to the risk perceptions. Analyses reveal that six AR measures were successful in predicting the perceived risk class of 23 companies (a success rate of 92 percent). Five of the measures: company size, dividend pay-out ratio, asset growth, leverage and current ratio are well-known in accounting literature. The sixth is based on expected growth in earnings (EEG); and forecasting earnings is a widely observed activity amongst analysts. Higher (lower) EEG is found associated with lower (higher) risk perceptions. The explanation offered is that amongst companies exposed to similar risks, higher (lower) EEG signals lower (higher) risks to analysts.; These findings, first, suggest that publicly available accounting numbers can be used by private investors for risk assessment, Second, they provide a potential new area of search for the information content of earnings. Tests of the information content of earnings, using conventional accounting models, have so far yielded disappointing results. The earnings variable used in these models is earnings surprise. The EEG/risk perceptions association identified in this study is robust; EEG may therefore provide a higher explanatory power for earnings that is more consistent with the importance postulated for earnings in economic theories of share valuation. |
