“…So the standard error from the residuals of the forecasting equation (13) However, one drawback to this measure of uncertainty is that it is backward-looking instead of forward-looking as the theory assumes. It is possible to overcome this problem typically using survey data about future expectations of business performance with respect to production, market prices, etc., as done for example in Guiso and Parigi (1999) using a survey of the Bank of Italy, or Driver and Moreton (1992) using the CBI annual survey. But also this proxy of uncertainty, although forward-looking, is not immune from critics, for example about the fact that survey data contain only qualitative information on the forecasted performance 41 , and this implies some degree of subjective discretion on how to translate this information in a quantitative measure of uncertainty; moreover not all firms on the market are obliged to respond to the survey, leaving out generally the smaller firms which can, in the aggregate, contribute to a great deal of investment; and finally it is not always guaranteed that the person involved in filling in the questionnaire is the one who has decision power inside the company, and therefore might not possess all the necessary information about the future strategies of the firm, its intended investment plans or production possibilities.…”