Suppose you need to write an empirical dissertation based on a dataset that contains the following information about some large company XYZ. First, it contains, for each year in the period 1980-2018, financial analysts’ average predictions at the start of the year about XYZ’s sales for that year as well as for the subsequent year. For example, the dataset would tell you that, say, at the start of 1990, the average prediction (where the average is taken across analysts) was that XYZ’s sales would be £x million in 1990 and £y million in 1991. Second, the dataset also contains XYZ’s actual sales for each year in the period 1980-2018.
a) What interesting question can you try to answer in your dissertation based on the available dataset? Assume that, to avoid overlap with a fellow student’s dissertation, the question cannot be specifically about whether analysts over- or underextrapolate based on past sales. (The question that I want you to state was discussed in the lectures, but not specifically for data related to sales.) [12 marks]
b) How would you go about answering this question? In particular: What variables will you compute? What possible connection(s) between these variables will you look at? [18 marks]