Bernd Skiera, and a couple of colleagues, has a paper that considers a relatively recent wave of research in marketing academia. They tackle event studies. These look at the impact of specified “events” on the value of a firm. Clearly marketers tend to concentrate on marketing events – e.g., name changes, product launches etc… The basic idea is to use assumptions about the stock market and how it assesses the future value of a firm to see the impact events happening today have. One has to assume that the stock price is relatively efficient, i.e. that the stock price is a good prediction of the future value of the firm given what is currently known. Given this we can see the impact of any event on the firm, i.e. new information, through changes in the stock price. For example, does the market value of the firm go up or down with the announcement of a name change? This value is assessed using the abnormal (unexpectedly high or unexpectedly low) returns to the company’s stock around the time of the announcement.
The authors discuss the fact that marketing research typically compares the abnormal returns based upon shareholder value. They suggest an often better alternative is to only consider the impact on the operating business. This excludes the leverage effect from the ratio of operating value to total value. Basically they want to exclude the financing and other non-operational decisions of the firm from the analysis. This is important because the leverage effect can dampen, heighten, or even reverse the observed effects of an event.
The authors illustrate their ideas. They then simulate some events to show how the difference between the abnormal returns metrics can occur depending upon whether one considers only the operating business or the entire business. They then discuss when the differences might matter. It is also worth noting that the authors revisit three published papers to show how their advice changes the papers. One great thing to note is that all the original authors kindly shared their data. This sort of collaboration is excellent to see.
The implications are interesting. They say that: “The choice of the dependent variable is particularly important when comparing marketing performance across firms.” (Skiera, Bayer and Scoler, 2017, page 43). They summarize their point: “Thus, the aim of this paper is to encourage researchers to consider their choices of dependent variables carefully and to use [operating business cumulative abnormal returns] where appropriate, instead of [shareholder value cumulative abnormal returns], or to report results for both dependent variables and then argue which one is most suitable for the problem at hand.” (Skiera, Bayer and Scholer, 2017, page 4). (A dependent variable being what you think is being impacted by choices, e.g., financial performance).
They simplify this statement to “… we feel that the choice of the dependent variable in marketing-related event studies warrants much more discussion than it currently receives.” (Skiera, Bayer and Scholer, 2017, page 4). I totally agree but would suggest that they can go broader and drop the “-related event studies part”. Marketers need to get serious about what we think we are trying to impact – we won’t all agree, for example on what performance is, but we should be willing to explain our positions. If we can’t justify our choice of dependent variable why would anyone care that something seems to change it?
Read: Bernd Skiera, Emanuel Bayer & Lisa Schöler (2017) What Should Be the Dependent Variable in Marketing Related Event Studies?, Forthcoming in IJRM