Affordable access to electronic news and social media have increased the propensity of people to browse abundant opinions expressed by others and get influenced by those opinions while taking related decisions. The degree of the uncertainty looming over the optimality of the decision and its associated stake influence the intensity of this inclination. The stock market is one example where the uncertainty is high and so are the stakes. Ordinary investors skim through freely available expert opinions and recommendations in the social media on buying or selling a stock without knowing how much those advices were worth. Interpretation and assessment of an opinion get complicated because it is expressed in a natural language, such as English, which is not easily amenable to an unambiguous quantification of the expressed opinion. This research proposes a novel method of quantifying unstructured textual opinions of stock market experts in a genetic algorithmic framework. It explores to what extent the stock price movements of some stocks are more in sync with expert recommendations compared to other stocks, and how contrasting the predictions induced by the recommendations of different experts are. Empirical studies have been performed with a large volume of publicly available stock market data and associated expert opinions expressed in various social media. The findings indicate the proposed method to be a credible way of treating opinions in the domain of stock markets. By using the method an investor can empower herself while treating social media information in accordance with its merit.