The boats that did not sail: Asset Price Volatility and Market Efficiency in a Natural Experiment

Financial markets are thought to be inefficient when they move too much relative to the arrival of information. How big is this inefficiency? In today's markets, this is difficult to determine because the arrival of information is hard to identify. In this paper, I present a natural experiment from history in which the flow of information was regularly interrupted for exogenous reasons. This allows me to study volatility in the absence of news, and to identify the degree of inefficiency. During the 18th century a number of English securities were traded on the Amsterdam exchange. Relevant information from England reached Amsterdam on mail boats. I reconstruct their arrival dates. When no mail boats arrived, virtually no other relevant information reached the Amsterdam market. I measure price volatility during periods with and without news. Even in the absence of new information, security prices moved significantly. Between 50 and 75% of overall volatility did not reflect the arrival of news. A significant fraction of this residual is driven by the incorporation of private information into prices. Once this is taken into account, 20 to 50% of the overall return variance is unexplained by information. This suggests that the Amsterdam market moved more than can be explained by the arrival of news but that the majority of price movements was still the result of efficient price discovery.