The switch from continuous to call auction trading in response to a large intraday price movement

Abstract

Some European exchanges (e.g. Euronext, Frankfurt and Madrid) make use of a mechanism to moderate price volatility that was proposed by Madhavan (1992) as preferable to a trading halt in times of market stress. It consists of a temporary switch from continuous to call auction trading in an individual security whenever its price moves beyond predetermined limits. This article studies whether this mechanism sharpens the information content of prices, dampens volatility and normalizes trading volume and intensity. Taking intraday data for the Madrid order driven continuous market, I find post switch improvements in the information content of prices and reductions in volatility, especially for thinly traded stocks. Trading volume and intensity peaked around auctions, but soon returned to preevent levels.

Acknowledgements

I am especially indebted to Ian-Charles Coleman for his constructive criticism and corrections. I also thank Lluís Brú, José M. Ortiz Repiso, José A. Pérez, Roberto Pascual, Mikel Tapia and seminar participants at the Universidad Carlos III (Madrid), the Bolsa de Madrid, the Simposio de Análisis Económico 2004 and the Foro de Finanzas 2004 for their constructive comments on previous versions of this article, circulated under the title ‘The Switch from Continuous to Periodic trading as a circuit breaker. Evidence from the Spanish Stock Market’. I thank the Spanish Stock Exchange for providing the database and Rafael Villar for research assistance. I am very grateful for partial financial support from the Spanish Stock Exchange in the form of the Best Research Paper on Stock Markets Prize (2004). I am also grateful for financial support from the Center for Economic and Financial Research (CIEF) of the Caixa Galicia Foundation, from the Xunta de Galicia (under contract PGIDIT06PXIB201002PR and INCITE09201042PR), and from research grant MTM2008-03010.

Notes

¹ See Galper (1999).

² For reviews of the trading halt literature, see Kyle (1988), France et al. (1994), Harris (1998) and Kim and Yang (2004).

³ For a more detailed account of the SSE trading system and trading rules, visit http://www.sbolsas.es.

⁴ The official rules provide for resolution of possible ties between different prices.

⁵ The static and dynamic price collars are also used to control opening and closing auctions. When the final virtual resolution price of the auction lies outside one of the collars (the static collar in the case of opening auctions, and either the static or the dynamic collar in that of closing auctions), the auction is extended for 2 min. However, no such extension of volatility auctions is possible.

⁶ Parallel analyses were also carried out for subsamples defined by whether the auction-resolving price was higher or lower than the price in the last transaction before the auction. No asymmetry was found between these two subsamples as regards the behaviour of price discovery, price volatility, bid-ask spread or trading volume and intensity (results available on request).

⁷ Unfortunately, it has not been possible to gain access to data allowing analysis of the progress of learning during volatility auctions.