A Theoretical Model and Empirical Analysis of Components of Spread in Over The Counter Exchange of India

Student Name: Jyothi G. Rao

High frequency transaction data from January 1995 to December 1995 of the Over The Counter Exchange of India (OTCEI) were studied in this thesis for analyzing its bid-ask spread. The spread was first analyzed using existing methodologies available in the literature with further statistical enhancements. Then a new theoretical model was developed for decomposing the bid-ask spread into its constituent cost components of order processing, inventory holding and adverse information, arising in any dealer market (not necessarily only in OTCEI), by extending the classical Stoll's financial model for this problem. This new model was then applied to the OTCEI data and the cost components were extracted and compared with the ones obtained from the existing methods. The OTCEI data did not exactly conform to the proposed theoretical model and thus some empirical fine tunings were necessary to accomplish this goal. This led to the development of a data driven empirical model, which was built around the skeleton of the proposed new theoretical financial model. The relative importance of the cost components in OTCEI were found to be similar using both new theoretical and empirical models, though their magnitude differed significantly. The existing models however painted a somewhat different picture. In particular, the order processing cost was found to be much lower using the new methodologies, which was attributed to the other two components under the existing methodologies. The reasons for failure of OTCEI and subsequent Dave committee recommendations however, could be put in better perspective, based on the findings of the newly developed methodologies than the existing ones.