資訊、交易與股票價格行為---台灣股票市場之實證研究

全文

(1)

(2)

(3) !"#$% # "&$'% "& (#)* + $), -,."/$ 0/$&,), !#.,1"$2"3#)*04)."', 567839::;<=>?;@;A>>;AA> BCD7:E F A: G A> HI :: F AE G J> H KLM7NOP BQRSTUVWXYZ[\]^

(4) !"#$%& '(%)*+

(5) ' (, -.!/012345678978'(:%)*+;89; 8'(5#$ <=0123> ?@01A34 B-.!/01C<=01C?@01 Abstract. This paper examines the stock price behavior in the trading and non-trading periods for stocks listed on the Taiwan Stock Exchange over 1971-96. The results indicate that the trading-time return variances are higher than the non-trading-time return variances especially for the larger trading-volume quintiles. This result is consistent with the private information hypothesis. Moreover, open-to-open return variances are higher than close-to-close return variances. Since both the opening and the closing transactions are conducted by the call auction procedure, the results are consistent with the trading halt hypothesis but not with the 1.

(6) trading mechanism hypothesis. . Keywords: return volatility, trading mechanism hypothesis, call auction method, dealership method, private information hypothesis.

(7) . The price behavior over trading and non-trading periods represents an important direction for understanding the stock price formation process. Two questions remain unanswered. First, empirical evidence is not unanimous concerning the pattern of price behavior in the trading- versus non-trading periods. Thus, results based on the Taiwan Stock Exchange would provide additional evidence to clarifying this empirical issue. Second, the institutional setting for the Taiwan Stock Exchange differs from that for several major stock exchanges in the world. In particular, the opening price and the clearing price in the trading period are all based on the periodic auction method. This feature provides a simple way to test the implications of the trading mechanism hypothesis against the trading halt hypothesis. . The result indicates that in general the trading-time return variance is larger than the non-trading-time return variance and that the open-to-open return variances are larger than the close-to-close return variances and that. Over the whole sample period 1971-96, the trading-time return variance,. σ ,. is. 0.019%. relative. to. the. 0.008%. for. the. non-trading-period return variance, σ , with a variance ratio of 2.33. Moreover, the open-to-open return variance, σ , is 0.035% versus the 0.027% for the close-to-close return variance, σ , with a variance ratio of 1.26. These variance ratios are statistically significant when the Levene 2.

(8) F test statistics are considered. Moreover, the result indicates that the autocorrelation between trading-time returns and the subsequent non-trading-time returns, ρ ı ı+ , is significantly positive. In contrast, the autocorrelation. between trading-time returns and the preceding non-trading-time returns, ρ ı ı is relatively smaller in magnitude and on average close to zero.. Over the whole sample period 1971-96, the autocorrelation is 0.30 between trading-time returns and the following non-trading-time returns but only 0.01 between trading-time returns and the preceding non-trading time returns. For the trading- versus non-trading return variances, the result of higher trading-time return variance is consistent with the prediction of the private information hypothesis but not with the prediction of the public information hypothesis. That is, the result suggests that the higher trading-time return variances are driven mainly by the revelation of private information when informed investors trade in the trading period. The result also suggests that public information is not the major factor for the higher trading-time return variance. Finally, the result of positive autocorrelation between trading-time returns and the following non-trading-time returns is inconsistent with the trading noise hypothesis neither. For the open-to-open and close-to-close return variances, the result of higher open-to-open return variance is consistent with the prediction of the trading halt hypothesis but not with the prediction of the trading mechanism hypothesis. Since the Taiwan Stock Exchange adopts the same call market procedure in both opening and closing, the higher open-to-open versus close-to-close return variances cannot be attributed to the selection of trading mechanism. Rather, the result is consistent with the prediction of the trading halt hypothesis.

(9) . 3.

(10) Amihud, Y. and H. Mendelson (1987), ‘Trading mechanisms and stock returns: An empirical investigation’, Journal of Finance, Vol. 42, pp. 533-53. ______ ______ (1988), ‘Liquidity, volatility and exchange automation’, Journal of Accounting, Auditing and Finance, Vol. 3, pp. 369-395. ______ ______ (1991), ‘Volatility, efficiency, and trading: Evidence from the Japanese stock market’, Journal of Finance, Vol. 46, pp. 1765-89. Black, F. (1986), ‘Noise’, Journal of Finance, Vol. 41, pp. 529-543. Brown, M.B. and A.B. Forsythe (1974), ‘Robust tests for the equality of variances’, Journal of the American Statistical Association, Vol. 69, pp. 364-67. Conover, W. J., M.E. Johnson, and M.M. Johnson (1981), ‘A comparative study of tests for homogeneity of variances, with applications to the outer continental shelf bidding data’, Technometrics, Vol. 23, 351-361. Draper, P. and K. Paudyal (1997), ‘Microstructure and seasonality in the UK equity market’, Journal of Business Finance & Accounting, Vol. 24, pp. 1177-1204. Fama, E.F. (1965), ‘The behavior of stock market prices’, Journal of Business, Vol. 38, pp. 34-105. French, K.R. and R. Roll (1986), ‘Stock price variances: The arrival of information and the reaction of traders’, Journal of Financial Economics, Vol. 17, pp. 5-26. Huang, Y.S. (1997), ‘The size anomaly on the Taiwan Stock Exchange’, Applied Economics Letters, Vol. 4, pp. 7-12. Kyle, A.S. (1985), ‘Continuous auctions and insider trading’, Econometrica, Vol. 53, pp. 1315-35. Levene, H. (1960), ‘Robust tests for equality of variances’, in Contributions to probability and statistics (ed.) Olkin, I., Standford University Press, Palo Alto, CA. Oldfield, G.S., Jr. and R.J. Rogalski (1980), ‘A theory of common stock returns over trading and non-trading periods’, Journal of Finance, Vol. 35, pp. 729-51.. 4.

(11)