Trading aggression when price limit hits are imminent: NARDL based intraday investigation of magnet effect

• Main Authors: Mohamad, Azhar, Sifat, Imtiaz Mohammad
• Format: Article
• Language: English; English; English
• Published: Elsevier 2018
• Subjects: HG4501 Stocks, investment, speculation
• Online Access: http://irep.iium.edu.my/67778/1/JBEF%20RnR%20NARDL%20Magnet%20Intraday.pdf; http://irep.iium.edu.my/67778/12/67778_Trading%20aggression%20when%20price%20limit%20hits%20are%20imminent%20SCOPUS.pdf; http://irep.iium.edu.my/67778/17/67778%20Trading%20aggression%20when%20price%20limit%20hits%20are%20imminent%20WOS.pdf; http://irep.iium.edu.my/67778/; https://www.sciencedirect.com/science/article/pii/S221463501830025X
• Institution: Universiti Islam Antarabangsa Malaysia

Utilizing an experimental Non-linear ARDL technique (NARDL), this paper tests an ex-ante hypothesized side-effect of financial market circuit breakers called the magnet effect. The hypothesis states that, in large price swing scenarios, circuit breakers (limits or halts), by their very existence, invite trading activities towards themselves in a way that the prophecy of the trigger is fulfilled. Most empirical works testing this effect hail from East Asian exchanges, which typically employ a tight price band. Our empirical venue, Bursa Malaysia, is a marked exception, sticking to a ±30% limit since 1989. Employing high-frequency (millisecond) proprietary intraday data from 2015 to 2017, we examine the magnet effect through order aggression and price velocity as the possibility of a limit draws closer. We find evidence of moderate magnet effect for most stocks, suggesting accelerated trading activities proportionate to likelihood of a limit-hit. The effect is more pronounced for lower limit stocks. Interestingly, several upper limit scenarios also exhibit the opposite of magnet effect: the repellent effect, suggesting investors recoil from trading when a limit-hit appears imminent. We discuss several regulatory, industry, and academic implications of our findings.