Research On Lag Component And Bias Component Of PB And Their Capacity To Forcast ROE

Return on equity (ROE) is one of the most important index in measuring theability to get profit. Investors focus on the future ROE when they consideringthe return on investment. People us ually use the external factors (the macroenvironment, industry environme nt, business environment) and internal factors(the enterprise internal resources, the enterprise ability, market co mpetitive ness)to calculate the input and output, and then use the input and output to forecastthe future ROE. The purpose of this paper is to find a improved method toforecast the future ROE through decomposed the price/book va lue ratio (PB)into the lag component and the bias component, to provide decis ion basis forinvestors.During this study, the price/book value ratio is decomposed into the la gcomponent and the bias component firstly. By lags, we mean that une xpectedeconomic gains (losses) are recognized in book value over time rather thanimmediate ly, so that the book-to-market ratio is temporarily lower (higher) tha nits mean (one in the absence of bias) but tends to its mean over time. By bias, wemean that book va lue is persistently higher (lower) than market va lue, so thatthe book-to-market ratio is persistently above (below) one. Then, calc ulate thatthe price/book value ratio has the negative correlation with the lag component,and the positive correlation with the bias component.In the calculate of the lag component and the bias component, thencomparison the price/book value ratio ability to forecast the future ROE and thedecomposed price/book value ratio ability to forecast the future ROE. Thispaper use SYS-GMM to do empirica l analys is because the data belongs todynamic panel. Considering the secondary market was establis hed no more then20years, the system has ma ny defects, so this paper selected Shangha i andShenzhe n300index component to analysis. Through the emp irical ana lys isfound that the price/book value ratio almost has no ability to forecast the futureROE and the lag component and the bias component has a more strongprediction ability.