Despite connectivity issues, regulatory hurdles and liquidity barriers, traders are eager to try out algorithms.
Electronic trading and the use of algorithms in Asian markets continue to grow despite regulatory hurdles and technological bottlenecks. According to Goldman Sachs, the value of client-directed trade orders executed through algorithms during 2007 more than tripled in Japanese equities and increased six times in the rest of Asia.
“There has been a significant increase in the adoption of algorithmic trading on the buy side because, when leveraged properly, it can lower the overall cost of trading and reduce information leakage while producing significant productivity gains,” says Todd Lopez, head of Pan-Asia electronic execution sales with Goldman Sachs.
Traders in Asia can appreciate the benefits that algorithmic trading has brought to other markets and at the same time can leapfrog over some of the early adopter issues that the US and Europe experienced, says Gabe Butler, Hong Kong-based sales director at Investment Trading Group (ITG).
“Asia’s traders have been much more willing to test algorithms, having seen the benefits they can deliver,” he says. “They are starting to use algorithms to trade the most liquid stocks and to trade large baskets of stocks. This way they can get comfortable with the algorithms before moving onto more widespread usage.”
Research firm Tabb Group estimates that the buy side in Asia is routing about 6 percent of its orders now via an electronic platform to brokers’ algorithms. However, when factoring in those firms that place orders over the phone with their brokers, who then enter the orders into algorithms, that estimate jumps up to 15 percent to 20 percent, according to Tabb Group.
Algorithmic trading technology vendors like ITG are expanding their operations in Asia to serve the growing demand. Smaller firms, such as hedge funds, typically do their algorithmic trading through brokers like Goldman Sachs.
Goldman Sachs offers multi-asset trading tools and single point of access to multiple products across 60 exchanges and 23 countries, including China, according to Lopez.
“Goldman Sachs provides algorithmic trading to customers such as hedge funds, index funds, program traders and statistical arbitragers. In Asia, the firm offers nine different core algorithmic trading strategies-compared with 11 strategies in the US and 17 in Europe. All allow for a significant level of customization,” he says.
But there are many idiosyncrasies that need to be taken into account when implementing algorithms in the region, Lopez says. These include unusual trading hours, complicated tick rules, and foreign investor identification requirements such as the Qualified Foreign Institutional Investor rules in effect in several countries.
“Asia is much more of a geographical concept than a homogenous market,” he says. “There is a unique set of market structure and regulatory issues for each country in the region. So it is critical that algorithms in Asia be calibrated accordingly.”
ITG’s Butler says that staffing is up more than 40 percent since the beginning of the year, much of it in technology-related jobs, with more to come. He agrees that a major obstacle to rolling out algorithms is the lack of uniformity between the various markets in Asia.
“You can’t just lift a product from one market and have it work in another,” he says. “We have to look at each market individually-Singapore is different from Hong Kong, which is different from Japan, which is different from Australia. That process requires a lot of technological expertise and a lot of local expertise.”
The algorithms have to be designed to take into account these differences, he says. “For example, a closing rotation algorithm might be appropriate for some markets such as Australia or Singapore, but not in Hong Kong where there is no closing auction.”
Providers of algorithms need to look at short sell rules, lot sizes, spreads, and minimum orders size for crosses, says Brook Teeter, head of advanced execution services sales with Credit Suisse.
“The best algorithms have proven to be incredibly useful by coupling the tremendous computing power available today with the intelligence and intuition of a seasoned trader,” says Nick McDonald, head of Asian equities at agency broker Instinet, which was acquired late last year by Japanese brokerage Nomura Holdings.
McDonald also says that because of the speed and sophistication of algorithms, they have enabled alternative liquidity pools, such as Instinet’s Japan Crossing and CBX, to thrive by existing in tandem with the traditional Asian exchanges.
“Having a first-rate, global technology team in place to develop the next generation of algorithms is obviously a must. But beyond that, there can be issues in applying the technology to markets with various inefficiencies, of which there are several in Asia,” Instinet’s McDonald says. “Fortunately, nearly all of the major Asian regulatory bodies understand the impediment that inefficient market structures pose and are working to rectify these issues.”
Customization will continue to be a trend for algorithms in Asia, says Goldman Sachs’ Lopez. “Customization is also becoming increasingly important on an individual customer level as well-we have worked with several of our clients in Asia to come up with highly customized strategies that look to achieve very specific objectives when taking into account the underlying fund strategy and short-term alpha assumptions.”
Users can now either tweak the parameters from the front end of their systems, or Goldman Sachs’ internal team of financial engineers can make back-end changes on behalf of customers. “We are working to figure out what makes the most sense for the users to meet their execution needs,” he says.
In response to user concerns regarding the increasing number of algorithms available, Goldman Sachs launched an “algorithm of algorithms,” called Navigator. Users enter the parameters of the trade into Navigator, which then makes recommendations on which algorithm and timing to use. Since it is a wrapper around the existing algorithms, users can interface with Navigator only, with the algorithmic decision made behind the scenes, Lopez says. Navigator can be accessed via Goldman Sachs’ RediPlus execution management system, via any major order management system, or via a proprietary system.
According to ITG’s Butler, another important issue when using algorithmic strategies is to ensure that users have appropriate training and education to understand each program’s capabilities or limitations. “The more advanced an algorithm in terms of customization and fine-tuning, the more sophisticated its performance, but likewise, the more understanding the user has to have of it,” Butler says.
Of course, direct market access (DMA) is key to developing high-speed trading and several Asian markets will have DMA capabilities in the coming year, which will further develop the base for algorithm usage in Asia, says ITG’s Butler. “DMA is emerging or being seriously considered by the local trading bodies in Malaysia, Thailand, India and Indonesia,” Butler added.
Bursa Malaysia Berhad announced earlier this year that by early 2008 the exchange will implement DMA, expected to reduce the latency of trading order executions from the current system’s five seconds per transaction to milliseconds.
“DMA is essential for the Malaysian capital market to remain competitive in the global investment arena,” said Bursa Malaysia Berhad CEO Dato’ Yusli Mohamed Yusoff earlier this year.
Asia, however, has only scratched the service of algorithmic trading. As spreads narrow and markets open their doors to algorithms-both technologically and through regulation-volumes will increase significantly, according to Credit Suisse’s Teeter.
Tabb Group also sees continued growth in algorithmic trading in Asia over the next several years. As more capital flows into the region, brokers will focus on pushing technologies that have a proven track record in the US and Europe into Asia, says Tabb Group senior analyst Adam Sussman.
“Algorithmic trading is here and likely here to stay,” said Gerald Greiner, deputy COO of Hong Kong Exchanges and Clearing, in a July newsletter. “We are in the process of considering a series of initiatives to support this business. These include enhanced market data services, direct market access, remote exchange participants, and facilitating hosting of participants’ technology. We are upgrading overall system capacity, which will increase speed-also known as reducing latency-which is very important to algorithmic traders,” Greiner said.
|Asia Markets Checklist|
|Characteristics||FIX adoption||Direct Market Access||Investor ID required||Limited short selling|
|Source: FIXGLOBAL, Securities and Exchange Board of India, FIX PROTOCOL|