Content area
Full Text
1. Introduction
Across the world's exchanges, researchers have observed a positive relationship between volume and volatility (e.g. [32] Lockwood and Linn, 1990; [33] McInish and Wood, 1990; [18] Foster and Viswanathan, 1990, [19] 1993; [22] Gerety and Mulherin, 1992; [2] Aggarwal and Gruca, 1993). The link has been shown robust to the country, traded asset and unit of time, and has motivated a theoretical literature that is central to market-microstructure theory[1]. Nonetheless, economists have only analyzed the volume-volatility link on exchanges that use continuous auctions to discover prices, despite the fact that there is evidence that market design matters[2].
Understanding agricultural commodity price discovery is an important and significant aspect of agricultural price analysis and to our understanding of how financial markets applied to commodity futures markets operate. The greater part of our understanding comes from open-outcry or electronic markets based on bid-ask spreads but there are alternative market forms that are not well understood and provide a unique glimpse into alternative forms of price discovery and market clearing. One of these is the Tokyo Grain Exchange (TGE), which finds prices for futures contracts using several daily sequences of discrete Walrasian tâtonnement auctions. A Walrasian tâtonnement auction on the TGE begins with a live auctioneer posting a tentative price. In response, traders submit pledges to buy or sell particular quantities if that price becomes the market price. If the tentative price does not clear the market, the auctioneer adjusts it in the direction of excess demand, and traders adjust their pledges given the new tentative price. This process continues until a tentative price is found that clears the market, at which time, the auction ends and all pledges are converted to actual trades at the market clearing price. The word tâtonnement is French for groping or trial-and-error.
Using six years of highly detailed data, we estimate the correlation between market activity and volatility on the TGE. The study is important for several reasons. First, since all previous studies have analyzed data from continuous auctions, analyzing the relationship using data generated from a completely different type of market microstructure yields insights into the extent to which the relationship is driven by information and the extent to which it is driven by market design. Second, compared...