A study of the pricing decision of new issues in Hong Kong 1970-1974.

  • Kim Y. S. Cham

Student thesis: Doctoral ThesisPhD


This study represents a new approach to the pricing decision of new issues. The past studies reveal two notable inadequacies, namely, the lack of prediction power and the failure to recognize the behavioral differences between individual issuing houses. Focusing on the market discount, Merrett, Howe and Newbould proposed the use of the tender method when pricing a new issue, thus avoiding the need for price prediction. Davis and Yeoman attempted to relate the market discounts to market conditions and financial ratios of the issuing company. The resulting low explanatory power precludes the application of the model to pricing decisions. Furthermore the use of only one model assumes an uniform pricing behaviour for all issuing houses; it is thus incapable of identifying individual pricing differences. In testing the market discounts of two issuing houses in the United States, McDonald and Fisher though recognizing the difference in their pricing behaviour, made no attempt to identify the variables leading to the pricing difference. The present study examines the actual pricing process of new issues during 1970 - 1974, taking into account individual behavioral differences. Pricing models have been constructed for nine issuing houses, of which six can be accepted unequivocally while the remaining three can be accepted with some qualifications. The results of the pricing models in terms of R2 are 0.55 for Wardley, 0.74 for Jardine Fleming, 0.79 for Schroders and Chartered, 0.69 for Hang Seng Bank, 0.81 for Oriental Financial Consultants, 0.94 for Union Bank, 0.92 for Hang Lung Bank, 0.67 for Overseas Trust Bank, and 0.50 for Bangkok Bank. All of them prove to have good predicting powers, even though the period of study included several volatile market conditions. In addition, individual pricing models were constructed for the four stock exchanges of the Hong Kong market, which yield satisfactory results. The R2 tests give 0.53 for Far East Exchange, 0.48 for Kam Ngan Stock Exchange, 0.52 for Hong Kong Stock Exchange, and 0.53 for Kowloon Stock Exchange. All in all, the models constructed in this study have demonstrated good explanatory and predicting power. In the Appendix, a supplementary approach has been used to construct models with a broad sample base but covering some unusual market conditions. The results obtained by using this approach, reveal that the approach adopted for this study is decidedly preferable in nine out of thirteen cases.
Date of Award1978
Original languageEnglish
Awarding Institution
  • University of Bath

Cite this