Quick introduction I am supposed to use a multiple linear regression model to determine the correlation between the market capitalization of a subset of european companies (primarily in Switzerland, Germany and Austria) and certain financial key data (Revenue, Total Assets, Equity...) - which exactly will be determined a bit later. I'm aware of the limitations regarding the utilization of backwards looking (financial statement) data, a change in the utilized modell or methodology isn't possible though, as identifying the limitations the multiple linear regression model has in this regard is part of the key findings that are being looked for.

The Problem(s)

Problem A: The Market capitalization at which Date should be used. Either the

  • Option 1: Market Capitalization 3 months after the release of the financial statements. (t+3)

This is based on the assumption, that the reported key values will have the strongest impact (if any) shortly after the year-end reporting of the fiscal year, which is released between 2.27 months after the end of the fiscal year in this set of data.

  • Option 2: Market Capitalization at the day of the financial statement (t)

Adversly it could be argued, that the quarterly reports before will already have set the expectations for the yearly financial statement, which could allow the utilization of the Market capitalization at the date of the financial statement

Primarily i tend towards Option A, as this has been widely used in the literature and past thesis I found in this regard. I might end up doing the same analysis for the market capitalization of both dates to analyze the difference. Do you think this would have merit?

Problem B: Especially in regards to Option 1 there is a certain problem I couldn't find an satisfying answer to in literature as most analysis in this regard have been done utilizing a single currency exclusively.

In my dataset I have both companies in EURO and in SWISS Franks. Obviously, both have to be converted into a single currency (primarily EURO in this regard).

However, the problem arises that, the market capitalization will be subject to the exchange Rate on the date (t+3) while the explaning variables will be subject to the exchange rate on the date (t), which will result in problematic deviations.

2 Solutions come to my mind:

  • Utilize the relative delta between the exchange rate T and the exchange Rate T+3 as a variable within the modell.
  • Use the exchange rate on date T to transform the market capitalization on (T+3) into EURO in order to disregard the exchange rate issue completely.

Neither solution seems a 100% clean so I wouldn't mind if one of you knows a better solution from personal experience. Otherwise I'd gladly hear your opinion on which one seems preferable to you.

I'd greatly appreciate your input on this topic, if there are any literature you could recommend looking into I'd greatly appreciate that as well as i seem unable to find much (especially in regards to the second problem). As mentioned before, utilized a different model is not possible as the multiple linear regression modell is part of the research question itself. (Limitations will be duely noted).

Some Literature to topic 1 includes:


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.