acm-header
Sign In

Communications of the ACM

ACM News

Computer-Driven High Speed Trading Improves Market Liquidity


View as: Print Mobile App Share:

Responding to concerns about an increasingly electronic stock exchange, Terrence Hendershott studied algorithmic trading and found that computer-driven trading based on algorithmic formulas improves the market's liquidity. Furthermore, this kind of high speed trading allows stock prices to become more "efficient" or reflective of true supply and demand in the market, says Hendershott, an associate professor at the University of California, Berkeley's Haas School of Business.

Forthcoming in the Journal of Finance, "Does Algorithmic Trading Improve Liquidity?" is co-authored by Charles M. Jones, Graduate School of Business, Columbia University and Albert J. Menkveld, VU University Amsterdam, Tinbergen Institute.

The research summarizes these key findings:

  • Algorithmic trading narrows the spread between the stock's bid and ask price.
  • It reduces adverse selection, which occurs when buyers and sellers make decisions based on a different set of information and results in both kinds of investors acting adversely based on the known, rather than the accurate, level of risk.
  • It reduces trade-related price discovery, meaning activity more truly reflects actual supply and demand.

The research data showed no evidence that algorithmic trading causes instability or volatility in prices. Hendershott and his colleagues spent two years gathering data at the New York Stock Exchange. In 2003, the exchange implemented a change in trading practices, speeding up how fast data was delivered to market participants. The upgrade and increased algorithmic trading were introduced across stocks over time, allowing later affected stocks to act as the research's control group.

"Does Algorithmic Trading Improve Liquidity?"  is available at http://faculty.haas.berkeley.edu/hender/Algo.pdf.


 

No entries found

Sign In for Full Access
» Forgot Password? » Create an ACM Web Account