Impact of low latency trading on market liquidity: We study the impact of low latency trading on market liquidity using a natural experiment at National Stock Exchange of India. We ...
Professor of Finance and Analytics at Great Lakes Institute of Management (India)
Impact of low latency trading on market liquidity: We study the impact of low latency trading on market liquidity using a natural experiment at National Stock Exchange of India. We find that introduction of co-location facilities leads to a significant improvement in market liquidity; quoted spreads and effective spreads decline across all stocks. Studying the impact on the components of effective spreads offers some new insights. While revenue to liquidity suppliers decreases, the cost of adverse selection increases; however, the former dominates the latter, leading to an overall compression in spreads. Time to complete execution of marketable orders – a key metric for traders seeking immediacy – also reduces significantly. Time to execution of passive non-marketable orders increases; their fill rates also decrease. These suggest that passive orders stay in the order book for a longer duration.
Low latency trading and the comovement of order flow, prices, and market conditions: We examine the impact of algorithmic trading (AT) in equities on the comovement of order flow, returns, liquidity, and volatility to assess how AT affects the market’s susceptibility to systemic shocks. Using order-level data around a natural experiment at the National Stock Exchange of India, which in 2010 has introduced features that promote HFT, we find that more intense AT reduces commonality in order flow, returns, liquidity, and volatility, and therefore reduce the market’s susceptibility to systemic shocks. These declines are more pronounced for algorithmic order flow and for large-cap firms. We attribute our findings to more intense competition among algorithmic than non-algorithmic traders.
Comovement of order-flow and liquidity in equivalent assets: We examine comovement in order-flows and liquidity of stocks and their single stock futures using a unique database from National Stock Exchange (NSE) of India that permits unambiguous identification of Algorithmic Trading (AT) activity. Order-flows and liquidity in both cash and futures markets are jointly determined. Order-flows in both markets are significant in explaining future returns. Using a natural experiment at NSE, namely introduction of co-location facilities, we find that the forecasting ability of order-flow generally decreases with reduction in latency. Immediacy demanding orders from non-AT traders have higher information content than those from AT. Order-flows in these markets are also jointly determined with mispricing in futures contracts. However, the impact of a transient order imbalance on mispricing has decreased significantly with the introduction of colocation facilities. We also examine comovement in unsigned order-flow and volatility of the underlying securities. We provide evidence of bidirectional causality between these variables. However, we do not find any single class of trader – AT or non-AT – dominating this relationship.
|Thesis Committee :||
Supervisor: Ekkehart Boehmer, EDHEC Business School
External reviewer: Kingsley Fong, Associate Head of the School of Banking & Finance, University of New South Wales
Other committee member: René Garcia, EDHEC Business School