The article presents the results of a research conducted on share buybacks done by companies listed on WSE in years 2005-2010. In most cases, the announcement of share repurchase program was designed to deliver costless information about stock underpricing. and was followed by rather sluggish repurchase actions. Next, I check the credibility of open market share repurchase programs by measuring the signaling effect among groups of small, medium and large companies. This is because of the assumption, that the credibility of companies declarations improves with company size. The event study methodology shows, that a share buyback announcement is a positive information for the shareholders and allows to obtain statistically significant abnormal returns the first day after the announcement. Abnormal returns’ characteristics such as fat tails and heteroskedasticity can cause the statistical inference bias that is based on the assumption that the errorprocess is homoskedastic and normally distributed. Because of this reason, I use a more robust approach, based on a two-stage bootstrap approach. Additionally, the companies were divided into companies with respect to high and low MV/BV ratio. The statistical analysis and calendar method shows long-term positive effect of abnormal share price increase for low MV/BV companies.