Post by Sapphire Capital on Jul 11, 2008 7:30:52 GMT 4
Market Makers versus the General Public: A First Look at S&P500 Futures Trade Data
GERARD L. GANNON
Deakin University - School of Accounting, Economics and Finance
--------------------------------------------------------------------------------
March 18, 2008
Abstract:
What has been undertaken in this research is a careful sampling of CFTC S&P500 futures trade records into the 15 minute required reporting intervals for the period January 1994 to June 2004. Accumulated volume of trade open, close, high and low prices are extracted for market trade and also market makers CT1 trading with the general public CT4 for each group selling short to the other. These trading records are matched with similar price records for the S&P500 cash index. An identifiable system of Simultaneous Volatility Model equations is artificially nested and tested, via a systems AIC, against a competing identifiable Structural VAR system. Results are reported from the dominant systems of Simultaneous Volatility Model equation estimates with futures trading volume, futures volatility and cash index volatility included as endogenous variables.
As we disaggregate from the market records to CT1 and CT4 records and further into year to year samples, volume to futures volatility leading effects and also futures volatility to cash volatility leading effects dominate. For the sub-period 1994-1999 for CT1 the leading volume term is significant for every year 1994, 1995, 1996, 1997, 1998 and 1999. For the latter sub-period the leading volume term is significant for 2002, 2003 and 2004. As with the annualized results for CT1, for CT4 estimates of the leading volume term are very significant in the futures volatility equation for all separate years 1994, 1995, 1996, 1997, 1998, 1999 and also 2002, 2003 and 2004. In the cash volatility equation for CT1 the lagged futures volatility estimate is very significant and lagged cash volatility insignificant for years 1994, 1995, 1996, 1997 and 1998 but not thereafter. For CT4 records the lagged futures volatility estimate is very significant and lagged cash volatility insignificant for years 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2002 and 2004. So although there appears to a deterioration in the aggregated data for the three groups as the analysis moves into 2000 the annual CT1 and CT4 results of volume and volatility lead/lag effects are quite strong.
The results raise important issues for risk management and dynamic hedging models employing intra-day trader data. A number of important issues for further analysis are also raised in this paper.
papers.ssrn.com/sol3/Delivery.cfm/SSRN_ID1108861_code483128.pdf?abstractid=1108861&mirid=4
GERARD L. GANNON
Deakin University - School of Accounting, Economics and Finance
--------------------------------------------------------------------------------
March 18, 2008
Abstract:
What has been undertaken in this research is a careful sampling of CFTC S&P500 futures trade records into the 15 minute required reporting intervals for the period January 1994 to June 2004. Accumulated volume of trade open, close, high and low prices are extracted for market trade and also market makers CT1 trading with the general public CT4 for each group selling short to the other. These trading records are matched with similar price records for the S&P500 cash index. An identifiable system of Simultaneous Volatility Model equations is artificially nested and tested, via a systems AIC, against a competing identifiable Structural VAR system. Results are reported from the dominant systems of Simultaneous Volatility Model equation estimates with futures trading volume, futures volatility and cash index volatility included as endogenous variables.
As we disaggregate from the market records to CT1 and CT4 records and further into year to year samples, volume to futures volatility leading effects and also futures volatility to cash volatility leading effects dominate. For the sub-period 1994-1999 for CT1 the leading volume term is significant for every year 1994, 1995, 1996, 1997, 1998 and 1999. For the latter sub-period the leading volume term is significant for 2002, 2003 and 2004. As with the annualized results for CT1, for CT4 estimates of the leading volume term are very significant in the futures volatility equation for all separate years 1994, 1995, 1996, 1997, 1998, 1999 and also 2002, 2003 and 2004. In the cash volatility equation for CT1 the lagged futures volatility estimate is very significant and lagged cash volatility insignificant for years 1994, 1995, 1996, 1997 and 1998 but not thereafter. For CT4 records the lagged futures volatility estimate is very significant and lagged cash volatility insignificant for years 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2002 and 2004. So although there appears to a deterioration in the aggregated data for the three groups as the analysis moves into 2000 the annual CT1 and CT4 results of volume and volatility lead/lag effects are quite strong.
The results raise important issues for risk management and dynamic hedging models employing intra-day trader data. A number of important issues for further analysis are also raised in this paper.
papers.ssrn.com/sol3/Delivery.cfm/SSRN_ID1108861_code483128.pdf?abstractid=1108861&mirid=4