This paper examines whether and to what extent announcements of the Federal Open Market Committee (FOMC) minutes have new and surprising information, albeit lagged, for the future expectation of monetary policy in the US. I construct new surprise series with intradaily data for the Fed future contracts and the responses of stock markets, fixed income markets and exchange rates to these surprises during 2004–2017. I found that the release of FOMC minutes affects the market volatility and financial asset prices respond significantly to FOMC minutes announcements. Finally, volatility and the volume of reactions increased during the zero lower bound. Specifically, this research finds that the release of FOMC minutes induces *'higher than normal'* volatility and shows that financial markets respond *'quickly and significantly'* to the release of FOMC minutes.
This research claimed that to evaluate the effects of monetary policy announcements correctly, a measure of the monetary policy surprises is needed. Specifically, as the efficient market hypothesis assumes that financial asset prices respond only to unexpected policy news or actions, this entails the necessity to measure surprise components. For example, if a monetary policy announcement, albeit contains a substantial change, is entirely expected, it will not have any surprise on the market; thus, financial asset prices will not change since the action will already have been priced in. In this regard, I found that the daily window is too long to fulfil this requirement and further, suggested the use of intra-daily data as an alternative, particularly for the analysis in recession times.