City Research Online

Essays on Cross-Sectional Asset Pricing

Lu, C. (2009). Essays on Cross-Sectional Asset Pricing. (Unpublished Doctoral thesis, Cass Business School)


The dissertation aims at the further understanding of several critical issues in the stock markets. It contains four chapters.

Cross-sectional stock returns and asset pricing has been one of the most important areas in financial economics. With the empirical failure of the Capital Asset Pricing Model (CAPM), an increasing number of studies have been conducted in the US stock market, and consequently many alternative asset pricing models and factors, have been proposed. Chapter One investigates the role of liquidity risk in cross-sectional asset pricing in both the USA and the UK. This study finds that a liquidity-augmented CAPM explains asset returns. Liquidity explains a sizeable spectrum of cross-sectional stock returns; and its effect is robust in the presence of other well-known empirical factors and a range of macroeconomic factors. Given the influential work of Fama and French (1992 and 1993), the performance of size and value premiums, (i.e., the excess return of small-capitalization stocks over large-capitalization stocks and the excess return of high book-to-market over low book-to-market stocks) are also compared. It is found that value premium is robust while the size premium disappears in the data for both countries.

Chapter Two investigates the relationship between liquidity and beta, as this relationship has been given little attention in the literature. Using the illiquidity measure of Amihud (2002), Acharya and Pedersen (2005) show that liquidity is priced in the framework of CAPM, and illiquid stocks have higher betas. This study, however, provides empirical evidence that Amihud’s measure is highly correlated with firm’s size, and the results of Acharya and Pedersen (2005) could be spurious because of inappropriate choice of liquidity proxy. Using the size-free liquidity measure proposed in this study, it is demonstrated that liquid stocks have higher betas. This is consistent with the model of Holden and Subrahmanyam (1996), in which risk-averse investors resist holding risky (high beta) stocks. As a consequence, they trade risky stocks more often than low beta stocks, thus increasing the liquidity of high beta stocks. The evidence that illiquid stocks have low betas while still commanding higher returns implies that liquidity is priced in a multifactor, rather than CAPM, framework, which is consistent with the work of Brennan and Subrahmanyam (1996) and Pastor and Stambaugh (2003).

In Chapter Three, many different factors proposed in the cross-sectional asset pricing literature are reviewed; and it is argued that the number of factors in the literature seems to be too large, as suggested by the Arbitrage Pricing Theory (APT). It is hypothesized that all the existing factors cannot be mutually exclusive and/or equally important, thus there must be redundant factors. More importantly, many of the successful factors are not well economically or theoretically motivated. For example, there is still no consensus on the underlying risk of the well-known Fama and French factors. Last but not least, many of these successful empirical factors suffer in terms of the data-mining critique of Lo and MacKinlay (1990). In this study, a total of 18 factors are assembled and categorized into three groups: five risk-related, eight firm characteristics and five APT motivated principal component factors. Individual stocks rather than portfolio returns are used in testing factor models to avoid the data-snooping problem. The results suggest a risk-related four-factor model can serve as a replacement for the controversial Fama-French and momentum factors. More importantly, the four factors, i.e., excess market return, co-skewness, downside risk, and liquidity, are economically and theoretically better motivated than the firm-characteristics based factors. It is also found that many of these firm-characteristics sorted factors are not pervasive in explaining individual stock returns. It is, therefore, concluded that most of the factors are redundant and may be the outcome of data-mining.

Chapter Four examines the cross-sectional effect of the nominal share price. This chapter endeavours to understand two interesting puzzles associated with share price. First, the nominal share prices of the US stocks have remained remarkably constant since the Great Depression despite inflation. Second, there is no consensus about the motivations for firms to split their stocks, since financial theory suggests share price is independent of its value. The findings indicate that share price per se matters in cross-sectional asset pricing: stock return is inversely related to its nominal price. It is shown that a strategy of buying these penny stocks can generate a significant alpha even after considering the transaction costs. The abnormal returns of these penny stocks are robust in the presence of other firm characteristics such as size, book-to-market equity, earning/price ratio, liquidity and past returns; and are also not explained by the existing factors. These results also cast some light on the stock-split phenomenon. Intuitively, if firm managers know that low price would generate higher future returns, they are more likely to split their stocks on behalf of shareholders.

This thesis makes several major contributions in the area of cross-sectional asset pricing. First, it highlights the importance of liquidity risk in the financial markets. For example, Chapter One and Three suggest the robust significance of liquidity risk in both the UK and US stock markets. Second, this study investigates the interaction between liquidity and other well-known factors in asset pricing. For instance, the well-documented value premium can be explained by liquidity risk (Chapter One), by the close link between liquidity and beta (Chapter Two); and by the close association between liquidity and size, share price and other factors (Chapters One to Four). Third, this study addresses the issue arising in the asset-pricing literature regarding the number of factors used in explaining asset returns. Chapter Three concludes that many of the existing empirical factors are not pervasive and may be the outcome of data-snooping as a result of grouping. Consequently, this chapter indicates that a theoretically better justified four-factor model, comprising excess market return, co-skewness, downside risk and liquidity, is competent to explain stock returns. Last but not least, this thesis also challenges the Efficient Market Hypothesis. Chapter Four demonstrates that investors buying low price stocks (penny stocks) and selling high price stocks can generate significant profits, and rational asset-pricing models cannot explain this abnormal return. Nevertheless the inverse relationship between share price and return does shed some light on stock split motivations.

The results of this thesis, suggest a number of future research projects. For example, most of the academic work on cross-sectional returns and asset pricing are accomplished for the major developed markets such as the UK and US. With the maturation and growing importance of emerging markets, it is feasible to test asset pricing hypotheses in these markets. The extent to which these hypotheses are validated in the emerging markets would significantly impact both academia and practitioners.

Publication Type: Thesis (Doctoral)
Subjects: H Social Sciences > HG Finance
Departments: Bayes Business School > Finance
Doctoral Theses
Bayes Business School > Bayes Business School Doctoral Theses
[thumbnail of Lu, Chensheng.pdf]
Text - Accepted Version
Download (3MB) | Preview


Add to AnyAdd to TwitterAdd to FacebookAdd to LinkedinAdd to PinterestAdd to Email


Downloads per month over past year

View more statistics

Actions (login required)

Admin Login Admin Login