Abstract
In this thesis, I assess the predictability of financial analysts' stock price recommendations through a case study of the Oslo Stock Exchange (OSE). The foundation of this assessment is ~23 000 recommendations published to the Institutional Brokers' Estimate System (I/B/E/S) between 1st of January 2007 and 31st of March 2017 associated with stocks composing the OBX Index. Using generalized least squares (GLS) techniques, I, firstly, document financial analysts' would-be raison d'être on the OSE: (i) significant positive (negative) multi-factor model adjusted excess return associated with 'Buy' ('Sell') recommendations and (ii) significant increases in share turnover, both at market and broker level, in days subsequent to publication. Secondly, I find indications that these returns are likely reaped by the recommendation issuer's clients and, in particular, institutional investors: (i) significant above-average bought (sold) value volume associated with 'Buy' ('Sell') recommendations and (ii) significant increase in trade size, a symptom of institutional investor activity, for the same broker that issued the recommendation in days following its publication. Lastly, and contrary to the bandwidth these recommendations are granted in Norwegian financial news, I find that they likely have little / no systematic value to the general public: e.g., no significant long-term multi-factor model adjusted excess return post-publication, in conformance with the Efficient Market Hypothesis.