Dark data. Some methodological issues in finance

Abstract

The nature of the data of financial systems raises several theoretical and methodological issues, which not only impact finance, but have also philosophical and methodological implications, viz. on the very notion of data. In this paper I will examine several features of financial data, especially stock markets data: These features pose serious challenges to the interpretation and employment of stock markets data, weakening the ‘myth of data’. In particular I will focus on two issues: (1) the way data are produced and shared, and (2) the way data are processed. The first raises an internal issue, while the second an external one. I will argue that the process of construction and employment of the stock markets data exemplifies how data are theoretical objects and that ‘raw data’ do not exist. Data are not light and ready-to-use objects, but have to be handled conceptually and technically very carefully and they are a kind of ‘dark matter’. Dark data, for the note.