A comparison of alternative cash flow and discount rate news proxies

This paper evaluates various methods of estimating firm-level Cash Flow and Discount Rate news. The standard VAR approach of Campbell (1991) and Voulteenaho (2002) is compared to two alternatives that use analysts' forecasts as proxies for market expectations: the ICC method of Chen et al. (2013) and the Revisions in Analysts' Forecasts (RAF) method. Methods are evaluated based on (i) the forecast accuracy of cash flow expectations, (ii) association with the other proxies of CF news, and (iii) Monte-Carlo simulations. The results suggest that analysts' forecasts behave as more accurate proxies of market expectations than the VAR system predictions. The ICC method is sensitive to the cash flow forecast errors because it amplifies forecast errors during implied cost of capital estimation. Overall, the RAF method appears to have a greater construct validity than the VAR and ICC methods as it incorporates forward looking information of direct cash flow forecasts and avoids amplification of forecast errors.