본문 바로가기
  • Home

Can the Cyber-Information Influence the Stock Price? : Using an Artificial Intelligent(AI) Algorithm Methodology

Mincheol Woo 1 Kim, Jee-Hyun 2

1한국거래소
2한림대학교

Candidate

ABSTRACT

This paper investigates whether the information in internet stock message boards can influence stock prices. Using the SVM(Support Vector Machine), which is one of artificial intelligent algorithm methodologies, we collect 7,054,316 messages posted on the three major internet stock message boards- Never, Daum, and Paxnet- for the period from September 2013 to June 2015. The total number of posting reads (hits) is 1,662,840,007. The major findings of this paper are as follows. First, we find significant evidence that the number of postings and the number of hits increase stock prices. Second, postings with bullish or bearish key-words have more significant effect on the stock prices. Third, these findings are more prominent with the KOSDAQ-listed stocks, suggesting that the possibility of unfair trading based on posting manipulation is higher with KOSDAQ-listed stocks. Forth, we find that the influence of the number of postings and hits are different according to stock price ranges, indicating that there exists various individual investors with respect to the use of messages on internet stock message boards.

Citation status

* References for papers published after 2023 are currently being built.