The Determinants of Common Stock Returns Volatility: An International Comparison

Abstract

IT HAS BEEN COMMONLY NOTED that the variance of stock market returns differs considerably from stock to stock, across exchanges, and among countries. This paper seeks to determine empirically the importance of "thinness" in explaining such differences. We essentially regard thinness as being inversely related to the size of the market for the equity shares of a particular corporation. More specifically, we call the market for a particular stock thin when its floating supply (number of shares outstanding less insider holding) and/or its price per share are relatively low.The issue of thinness is of interest for a number of reasons. To the extent that shareholders do not diversify their portfolios so as to effectively eliminate all diversifiable risk, thinness in the markets for individual issues can increase the cost of capital for individual firms. This suggests that, if thinness contributes to returns variance, a firm will realize a financial economy of scale as it becomes larger. The undesirability of thinness-induced variance for risk averse investors further suggests that investors will select securities with reference to thinness, as well as to inherent investment profitability. Also, to the extent that thinness adds to variance, differences among investors in liquidity needs are expected to generate a "clientele" effect.Factors in addition to market thinness also contribute to returns variance. The most obvious are changes in the fundamental determinants of share price (e.g., expectations of earnings per share) and of a firm's business and financial risk. We attempt to account for this by distinguishing between random tenders-induced demand shifts (which are thinness related) and demand shifts induced by the receipt of new and generally available information concerning a stock's value (which are not thinness related). We then show that, with the thinness measures held constant, a security's turnover ratio (TOR), the number of shares traded divided by the stock's floating supply (FS), can be used to proxy the receipt of new information.We also note that across exchanges (and especially internationally) differences in trading arrangements might explain some of the volatility differences we have * The authors' affiliations are: Cohen,

Liquidity and Price Variability in Futures Markets

“…The liquidity hypothesis rests on findings in the stock market that thinly traded securities exhibit higher variability than liquid securities [5,6,91. Other studies [7, 8,9,131 conclude that a reduction in trading activity will increase the bid-asked spreads and consequently increase the volatility of such spreads.…”

Os pesquisadores, as publicações e os periódicos da área de Finanças no Brasil: Uma análise com base em currículos da plataforma Lattes

“…(2002)). Salienta-se, porém, que se considerarmos artigos anteriores a 1990, encontramos outros dois autores brasileiros e três artigos neste mesmo periódico, conforme destacado por Leal et al (2003): Walter L. Ness Jr (Cohen et al, 1976) e Ney Brito (Brito, 1977(Brito, , 1978 Ao analisar o número médio de autores por artigo em função dos anos, painel (c) da Figura 3, observa-se que a quantidade média de autores aumentou de aproximadamente 2,15 autores em 2000 para 2,5 em 2014. Este resultado também foi encontrado em Leal et al (2013) e pode ser explicado pe-los benefícios para a contagem de pontos dos programas de pós-graduação pela CAPES, de forma a incentivar trabalhos com mais de um autor e entre instituições diferentes.…”

The Returns Generation Process, Returns Variance, and the Effect of Thinness in Securities Markets

“…We model the generation of both aggregate shifts and idiosyncratic tenders using mutually independent compound Poisson processes. 9 We assume that aggregate demand shifts are generated by a Poisson process with mean rate -y per unit time. We define a Bernoulli random variable W to indicate whether an aggregate demand shift is: (a) upward (i.e., increases the quoted price), in which case W= 1, which occurs with probability p; or (b) downward (i.e., decreases the quoted price), in which case W= -1, which occurs with probability 1 -p. When p = 0.5, there is no drift in prices over time; when p >0.5, there is upward drift (which ex ante we regard as typical for most common stocks which do not pay out all of their earnings as dividends); when p<0.5, there is downward drift in prices.…”