EOD prices are often used to determine the expiration value of directors' options, the price of seasoned equity issues, evaluate broker performance, calculate net asset values of mutual funds, and compute stock indices.

"On the one hand there's clear incentive to manipulate the closing price by ramping end of day trading to push the closing price to an artificial level," says Professor Michael Aitken, CMCRC CEO. "However, EOD dislocation could also simply reflect price pressure brought on by the fact that the market is about to close for 18 hours. Either way, EOD dislocation of prices is not a good look for markets."

The researchers examined data from 22 exchanges from around the world, from 2003-2011. They found that the presence of HFT decreases the probability of end-of-day (EOD) dislocation by 21%.

The study also examined specific dates when EOD price dislocation was most likely to be manipulation, including dates when options expire and end of month/quarter calendar dates. The data showed that in the presence of HFT, EOD price dislocation was less pronounced on these dates as well.

Aitken says that policy mechanisms, including trading rules, surveillance and enforcement, appeared to have had less of an effect in mitigating EOD price dislocation than HFT, suggesting that the market may indeed be capable of disciplining itself.

He says: "There is an established negative relationship between liquidity and EOD prices (i.e. the higher the liquidity the harder it is to manipulate or the less prices will move) and that HFT by either providing additional liquidity at these points (or because market participants know that HFT are present in a marketplace) appears to reduce EOD dislocation, whether it is caused by fair means or foul."

Data and methodology available for download: