•  
• Tweet
•  
•  
•        Print
•  
•        Email
•  
•         Track Keywords

The research by CMCRC PhD candidate Tony Zhang and Dr Vito Mollica from the MGSM showed that companies can influence the level of algorithmic trading in their own stocks, should they want to do so, by adjusting their share price level.

“Our data suggests that firms with higher relative tick sizes attract less HFT, perhaps due to the increased marginal cost of getting ahead in the order book,” he said. “We analysed data from 2009-2012 and found that when companies undergo reverse stock splits which lower their relative tick size, algo trading increased and remained elevated. The reverse was true for stock splits.”

Researchers compared the data to stock splits and consolidations from 1996-2004, before algorithmic trading had reached appreciable levels, and did not find the same effects. “This confirmed for us the conclusion that algorithmic traders were finding more reward in trading stock with lower tick sizes.”

Institutional investors have been vocal in recent years in condemning the complexity and speed of contemporary equity markets, saying that algorithmic trading and high frequency trading were filling order books with the ‘noise’ of excessive messaging traffic and cancellations. Many have called for higher regulated tick sizes which they say would favour long term investors and encourage investor confidence.

“Companies themselves have complained about the volatility of their stocks, and while it’s debatable whether algorithmic trading can be blamed for that, there is certainly action that companies can take to reduce the level of this type of trading in their stocks, if that’s something they want to do,” Dr Mollica said.

The study investigates two algorithmic proxies, order-to-trade ratio and order resting time, to see how they change when a stock’s relative tick size changes after a stock capitalisation event. The study uses evidence from 131 stock split/consolidations events between 1996 and 2012, where the relative tick size at least halved or doubled. The study finds during the algorithmic trading active period (2005-2012) that the order-to-trade ratio (calculated by dividing the total number of orders traded, deleted or amended by total dollar turnover) increased significantly when a share moves from a large relative tick size to a smaller one. A large order-to-trade ratio means that more orders are required to complete a dollar value trade. Similarly order resting time (the time each order stays on the system before it is traded or deleted) is reduced as a firm moves to a lower relative tick level.
km