¹ See Scott, I., The Government Securities Market. (New York: McGraw-Hill, 1968), p. 93.Google Scholar

² See Douglas, G. W., “Risk in Equity Markets: An Empirical Appraisal of Market Efficiency,” Yale Economic Essays (1969), pp. 3–45.Google Scholar

³ Stigler, G. J., “Public Regulation of the Securities Market,” Journal of Business (April 1964), p. 127Google Scholar. Douglas, “Risk in Equity Markets,” makes a similar observation.

⁴ Demsetz, H., “The Cost of Transacting,” QJE (February 1968)CrossRefGoogle Scholar; Tanner, J. and Kochin, L., “The Determinants of Differences between Bid-Ask Spreads on Government Bonds,” Journal of Business (October 1971)CrossRefGoogle Scholar; Tinic, S. M., “The Economics of Liquidity Services,” QJE (February 1972)CrossRefGoogle Scholar; Tinic, S. M. and West, R. R., “Marketability of Common Stocks in Canada and the U.S.A.: A Comparison of Agent Versus Dealer Dominated Markets,” Journal of Finance (June 1974)CrossRefGoogle Scholar; and West, R. R. and Tinic, S. M., The Economics of the Stock Market (Praeger Publishers, 1970).Google Scholar

⁵ Smidt, Seymour, “Which Road to an Efficient Stock Market,” Financial Analysts Journal (September/October 1971)CrossRefGoogle Scholar, and Barnea, Amir, “Performance Evaluation of New York Stock Exchange Specialists,” Journal of Financial and Quantitative Analysis, September 1974.CrossRefGoogle Scholar

⁶ Stigler, “Public Regulation,” p. 127.

⁷ See Sarnat, M., The Development of the Securities Market in Israel (Basel: Kyklos-Verlag, 1966)Google Scholar, Chapter 4. The trading procedures described here are no longer used on the TASE.

⁸ It is not clear why the official dealer does not trade for his own account. The only possible explanation is that there are a large number of institutional members of the exchange and they perform the same type of market-making function as the specialist on the New York and American Exchanges (see disucssion in text below).

⁹ Many of the securities on the TASE are in bearer form so that data on the number of investors could not be collected in any simple way. A sample of securities on the New York and American Stock Exchanges was examined with regard to the relationship between TA, DS, and number of stockholders. The correlation coefficient was .85 between each pair of these three variables.

¹⁰ See Theil, Henri, “Specification Errors and the Estimation of Economic Relationships,” International Statistical Institute Review, Volume 25, No. 1–3, pp. 41–51.CrossRefGoogle Scholar

¹¹ See Malkiel, B. G., The Term Structure of Interest Rates (Princeton, N.J.: Princeton University Press, 1966).Google Scholar

¹² Given that the raw correlation coefficients between DS, V, and P are above .8, there is likely to be a serious multicollinearity problem in the actual regression equations. This would have been true in any case because equation (2) would have had to include DS and P as explanatory variables in order to avoid constraining their coefficients to unity.

¹³ Douglas, “Risk in Equity Markets,” p. 33. He argues that low-priced securities are traded more heavily by speculators whose activity is likely to be destabilizing. On the other hand, the work by Heins, and Allison, , “Some Factors Affecting Stock Price Variability,” Journal of Business (January 1966) does not validate this hypothesis.CrossRefGoogle Scholar

¹⁴ See Appendix A for a more detailed description of the data and their sources.

¹⁵ Those variables which could take on a value of zero, e.g., DS, AP, and I were increased by .0001 so that the logarithm could be calculated. This preserved the ordering of the observations, that is, a large quantum jump from zero to the smallest observation (which was .01) with only a slight smoothing of the series.