Banks are pouring capital into developing dark pools in Asia to provide clients with ever faster and more cost-effective trading, and that spending is having the desired effect, according to research released yesterday by US-based dark pool provider Tora Trading.

“A couple of brokers in this report that were in the third or fourth quartile last quarter are now in the top quartile [in terms of the performance of their algorithms],” says Ian Lombard, chief operating officer at Tora in San Francisco. “They are both firms that have made massive improvements and spending on their headcount and systems.”

The study covered January to June and analysed trade data from more than 15 brokers across Hong Kong, Japan and Singapore processed through the Tora Compass trading platform.

Speaking to AsianInvestor during a trip to Hong Kong yesterday, Lombard would not disclose the names of any of the brokers analysed. And it could be difficult for an outsider to guess, as so many firms are investing in dark pools – off-exchange, anonymous crossing engines.

“Clearly some brokers are investing heavily and some not so much,” says Lombard, “and as a client that should factor into your decision of which to use.” Yet not all do – strong algo performance is not necessarily reflected by strong volumes traded through said algorithm, he says.

Barclays Capital and Morgan Stanley have been ramping up in this area, Deutsche Bank launched a Hong Kong dark pool this week, as did agency broker ITG in March. And that's not to mention the plethora of well established platforms in the market already.

“Companies are waking up to the fact they’re not too late to get in; there are still big benefits to provide to clients,” says Lombard. “We're seeing a number of brokers increasing investment in this space; all the main brokers either have it, are on the cusp of offering it or are talking about getting in.”

Lombard says Tora has been carrying out quarterly research on transaction performance for two years, and it shows that overall performance is improving, but not for all brokers – some are performing worse than they did last year.

The difference between the best and worst brokers – in terms of slippage, spreads, speed and cost of trading – is very small, but big enough to make a difference to those trading frequently, he adds. Tora’s research aims to help clients make better trading decisions by showing which are the cheapest venues.

The research also found wide variations in performance across markets, with Hong Kong and Singapore faring weaker than Japan. Average slippage was 15.1 basis points in Singapore, 9.5bp in Hong Kong and 4.5bp in Japan. Slippage is the difference between the intended price of a trade and the price where the trade is actually executed.

The study reviewed around 120,000 orders and showed performance varied by as much as 8bp across brokers in Asia and that algorithms, despite variations in performance, can offer cost savings compared to a traditional execution. Considering the frequency of trades executed by users of dark pools, 8bp can soon add up to significant cost savings.

“Algorithms are an important tool increasingly used by traders in Asia, yet the study results clearly show there are differences in performance,” said Robert Dykes, Tora’s chief executive. “The overall advantages are clear as long as traders are diligent about which broker's algo they use given their trading objectives.”

The analysis examined a range of factors that affect performance, including liquidity, volatility, spreads and user-defined parameters. “The key to maximising algorithm performance is to use the one that best reacts to prevailing market conditions,” says Dykes.