Guanfa Securities (GF Securities), one of the largest securities brokers in China, has selected StreamBase, in partnership with Thomson Reuters, to build their algorithmic strategies and high frequency trading (HFT) practice.
Last week, GF Securities announced they were building out a new trading system based on StreamBase’s complex event processing (CEP) platform. GF Securities’ new trading system is expected to be the fastest in China and will include pairs trading and order execution algorithms. The new trading system is end-to-end and is notable for the fact that as StreamBase is a full development platform for real-time events and keeps some of its software open source, GF Securities is able to able build a customised trading system to meet local-market quirks using only a small in-house team. SteamBase’s CEO, Mark Palmer in Boston, US, commented that this customisation was likely one of the reasons his firm’s software was selected, highlighting that all markets have some level of idiosyncrasies in microstructure.
Palmer noted that the deal was part of an overall trend in China and other Asian geographies, as local support for algos and HFT grows. In particular he noted the firm was seeing most interest in China, closely followed by India. He also saw some interest coming from Japan, though the market has traditionally been naturally conservative about making the move into algos and HTF, despite having very low latency. Moreover, interest was coming from Hong Kong as international players already customers of StreamBase in the US or Europe roll out strategies to key regional hubs.
Chinese and Indian brokers seem unfazed by recent trading crashes, with Palmer describing the attitude in both countries as very “entrepreneurial” as compared to, for example, Japan’s conservatism. He adds that due to the impact of the financial crisis, firms installing such systems were much more aware of risk and surveillance than European and US firms were at the time they were making similar upgrades. “It’s an advantage. They have better testing and infrastructure because this is now burned into everybody’s psyche,” he said.
On why the interest in Asia has risen so rapidly, Palmer commented that, “As with most markets, players have to compete. Early adopters have been moving into the US and bringing technologies back over to China.” Palmer cited CITIC’s recent adoption of Progress’s Apama software as just one example. Investor demand in China perhaps has some way still to develop, as Palmer noted he hadn’t yet seen algo products being used on a large scale for retail investors, but he anticipated brokers would begin to develop in this direction. In terms of speed, China was not at the level of “ultra-low latency yet”, he added Palmer. “It’s really where the US or Europe was eight to ten years ago… but it is changing rapidly.”
According to Aite Group, algos make up 24% of the volumes of equity traded in Asia in 2012, and will reach 31% by 2014. By comparison, the US currently has 64% or trading in algos, and will reach just 65% by 2014.
By Harry Thompson