ABSTRACT
This study analysed media agenda-building and intermedia agenda-setting in the context of IPOs, a news setting that is considered highly routinised and standardised, to shed light on both the direct and indirect influence of financial information subsidies on the news. Automatic content analysis was used to record the salience of different IPO issues in the press releases for IPOs listed on the main board of the Hong Kong Stock Exchange, and in the subsequent news reports. The relationships among the frequency counts of IPO issues were analysed using Poisson regression and zero-inflated Poisson regression to provide a more accurate understanding of the effects of press releases on the financial news via two pathways: agenda-building from source to media, and intermedia agenda-setting among news media at different stages of the IPO marketing period. The results provide evidence that press releases influenced the overall news agenda during the early stage of the offering period of the new stocks, although for some issues, negative or no relationships were found. In addition, issue salience in the news media during the early stage of IPOs showed significant transfer to the latter stages.
Acknowledgements
The work in this paper was funded by Research Grants Council, University Grants Committee of the Hong Kong Special Administrative Region, China (UGC/FDS14/H17/17).
Disclosure Statement
No potential conflict of interest was reported by the author(s).
Notes
1 This study examines all 14 Chinese newspapers in Hong Kong, including the Hong Kong Economic Times, Hong Kong Economic Journal, Apple Daily, Oriental Daily News, Ming Pao Daily News, Sing Tao Daily, Sing Pao, Ta Kung Pao, Hong Kong Commercial Daily, Wen Wei Po, Metro Daily, am730, Headline Daily and Sky Post.
2 The research team members included a veteran financial journalist and a PR practitioner. Other financial communicators were also consulted to finalise the list of issues.
3 To precisely record the salience of an issue and avoid repetition of keywords during a single mention of an issue, the program counted only one mention of each issue per paragraph. Therefore, in a press release containing 10 paragraphs, the count frequency was limited to a maximum of 10.