Empirical analysis of illiquidity premiums, their structural breaks, and predictive capability

Abstract

Both modern portfolio theory and asset pricing models assume a perfect capital market with unlimited liquidity. In practice, capital market turmoils such as the stock market crash of 1987 or the global financial crisis (GFC) of 2008/2009 illustrate the relevance of liquidity for asset pricing. As a result, an extensive research literature has developed, arguing that “liquidity” is a complex concept that cannot be observed directly. Nevertheless, the prevailing view is that investors are compensated with higher returns for less liquid assets. However, many studies show strong idiosyncratic behavior of illiquidity proxies, especially in the case of country-specific political or economic events. This dissertation consists of three studies analyzing illiquidity effects in capital markets. First, we provide a general introduction to the topic of illiquidity in chapter 1, before giving an overview of the superordinate research approach in chapter 2 and classifies the individual scientific works according to their thematic. Furthermore, we outline the main questions, data and methods as well as findings of these studies. Chapter 3.1 studies the relationship between the illiquidity of German real estate securities and their expected return for the period 2003 to 2017 using six different illiquidity measures. The observation period thus covers the global financial crisis of 2008/2009 and considers securities that are to some extent specific to Germany, namely open- and closed-end real estate funds. Overall, we find that market liquidity affects the market returns of our sub-samples throughout the sample period. In contrast to the results of Amihud (2002), these effects on the expected and unexpected illiquidity of the sub-segments are not fully consistent in terms of significance and sign. Rather, our results suggest structural break effects for the GFC period that require further investigation. We therefore examine the relationship between returns and illiquidity for structural break effects in Chapter 3.2. For better comparability, we use a broad sample of stocks from Germany (CDAX), the United Kingdom (FTSE All Shares) and the United States (CRSP) for the period 1999 to 2022. If we neglect structural breaks in the time series, we can confirm the literature. According to this literature, illiquidity in the capital markets decreases continuously and leads to insignificant analytical results for investors. Following the approach of Bai and Perron (2003), we obtain differentiated results for the three economies analyzed. The relationship between stock market excess returns and illiquidity is significant in many cases. Moreover, the signs of the coefficients on expected illiquidity change between crisis and non-crisis periods. When we add Fama-French factors, illiquidity of both the UK and the US samples loses their significance for the more recent periods. Since periods of capital market illiquidity are a good indicator of real economic and financial developments, we analyze the forecasting quality of a large number of economic parameters in Chapter 3.3. We find that the OIS-Spread, the financial market stress measure (CISS) and the bond credit spread (DEF) are particularly good indicators. The model extends the literature by considering a more recent sample from 1999 to the end of 2022, by including capital markets in Germany and the UK in addition to the US, and, in particular by analyzing the effect of current economic variables. Chapter 4 summarizes the research results and provides an outlook on open research questions.