Bid-ask spread, liquidity and the effects of firm-level and market-level features

Market liquidity is defined as the ability to trade quickly, with low transactions costs, at any time and with no or a minimal impact on price. Liquidity is crucial for well functioning of markets. It has many implications for traders, investors, exchanges, regulators, and the listed firms. Thus, an accurate understanding of liquidity concepts and its determinants is essential. This dissertation consists of three main sections relating to liquidity. We start with a comprehensive review of the frameworks currently available for understanding definitions and determinants of liquidity. We put a special emphasis on various liquidity measures discussed in the literature about equity markets. Indeed, measures that we present have specific properties and capture certain aspects of liquidity. Our purpose, however, is to highlight the differences and similarities of these measures and produce a more complete understanding of their limitations and extensions. To do this, we review and categorize virtually all the equity market liquidity measures. Secondly, we empirically evaluate the performance of five different methods appearing in the market microstructure literature in predicting the cost dimension of market liquidity, in other words "bid-ask spread". Microstructure literature proposes models that attempt to estimate bid-ask spread using low-frequency data. However, the question is whether low-frequency spread proxies really measure what researchers want to measure. This questioning is essential since inaccurate estimates of spreads can create misleading information about actual market liquidity and functioning of financial markets. Thus, we investigate the performance of these proxies on index, currency and gold futures trading in Borsa Istanbul Futures and Options Market (VIOP). Although market liquidity is widely investigated from the markets perspective, the effects of corporate level features mostly are neglected. To fill this gap, in the last section, we develop a log-linear model that combines both corporate-level and market-level aspects in determining the market liquidity of stocks worldwide. The results indicate highly significant effects of the selected features and yield policy implications.