• Refine Query
  • Source
  • Publication year
  • to
  • Language
  • 1
  • Tagged with
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • About
  • The Global ETD Search service is a free service for researchers to find electronic theses and dissertations. This service is provided by the Networked Digital Library of Theses and Dissertations.
    Our metadata is collected from universities around the world. If you manage a university/consortium/country archive and want to be added, details can be found on the NDLTD website.
1

Studies On Some Aspects Of Liquidity Of Stocks : Limit Order Executions In The Indian Stock Market

Chatterjee, Devlina 09 1900 (has links) (PDF)
We study some aspects of liquidity of stocks traded through the National Stock Exchange (NSE) of India. Initially we examine the multi-dimensional nature of liquidity by conducting day-wise factor analysis of eleven liquidity proxies across a cross-section of stocks, using data from two periods reflecting different market conditions. Five factors emerge consistently, interpretable as depth, spread, volume, price elasticity and relative activity. Subsequently, we study execution of limit orders in the NSE from three angles. First we consider order execution probability, using 106 stock-specific logistic models. Important predictors of order execution probability are price premium followed by volatility, relative activity, bid ask spread and order imbalance. Some differences are noted when comparing companies of different sizes and between buy and sell orders. Second, we study order execution times using survival analysis. Several diagnostic tests indicate that parametric Accelerated Failure Time models using the log-logistic distribution for the survival time S(t) are suitable for current data. 100 stock-specific models are built; results are consistent with the logistic models. Additionally depth is also found to be important. Finally we build 4 combined models across stocks for both execution probabilities as well as times. These models perform well on out of sample data, suggesting their predictive utility.

Page generated in 0.1023 seconds