Positional Momentum and Liquidity Portfolio Management

This thesis introduces a new positional momentum management strategy based on the expected future ranks of asset returns and trade volume changes predicted by a bivariate Vector Autoregressive (VAR) model. Chapter one provides some facts about the relationship between return and trade volume changes and the way they have been computed in general. It begins by investigating the simple VAR model to see if we can use the past values of return and trade volume changes to predict their current values. Then recent developments in portfolio management research on momentum portfolios are discussed. Chapter two introduces a new method to build a positional momentum and liquidity portfolios based on the expected future ranks of asset returns and their trade volume changes. This method is applied to a data set of 1330 stocks traded on the NASDAQ between 2008 and 2016. It is shown that return ranks are correlated with their own past values, and the current and past ranks of trade volume changes. This result leads to a new expected positional momentum strategy providing portfolios of predicted winners, conditional on past ranks of returns and volume changes. This approach further extends to a new expected positional liquid strategy providing portfolios of predicted liquid stocks. The expected liquid positional strategy selects portfolios of stocks with the strongest realized or predicted increase in trading volume. These new positional management strategies outperform the standard momentum strategies and the equally weighted portfolio in terms of average returns and Sharpe ratio. Chapter three introduces new positional investment strategies that maximize investors positional utility from holding assets with high expected future return and liquidity ranks. The optimal allocation vectors provide new investment strategies, such as the optimal positional momentum portfolio, the optimal liquid portfolio and the optimal mixed portfolio that combines high return and liquidity ranks. The future ranks are predicted from a bivariate panel VAR model with time varying autoregressive parameters. We show that there exists a simple linear relationship between the time varying autoregressive parameters of the VAR model and the autoand cross-correlations at lag one of the return and volume change series of the SPDR. Therefore the autoregressive VAR parameters can be easily updated at each time, which simplifies the implementation of the proposed strategies. The new optimal allocation portfolios are shown to perform well in practice, both in terms of returns and liquidity.