I find that the optimal price of a bet for a risk-averse bookmaker is a function of elasticity of demand and the number of outcomes of the betting event. In the presence of shocks to the order flow, however, the optimal price can change, and large adjustments can create arbitrage opportunities for informed investors. Using a large sample of online bookmakers and a unique data set of real-time betting odds, I find strong support for these predictions. Overall, the results shed new light on the efficiency of online betting prices. (c) 2021 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ ) Betting markets have experienced an unprecedented growth over the past few years (see, e.g., Mainelli and Dibb, 2004 for a comprehensive review). Online gambling, along with extensive deregulation and the abolition of national monopolies, has rapidly transformed betting from a segmented local affair to a sizable global market ( Vlastakis et al., 2009 ). 1 As a result, bookmakers compete with each other independently of where they are physically located. This shift from a monopolistic business to a highly competitive industry offers a unique opportunity to improve our understanding of how betting markets work. In this paper, I shed light on three properties of online betting prices. First, I analyze the determinants of bookmakers'