Published On: Fri, Nov 13th, 2009

High Frequency Trading Discussion in Hong Kong

Thomson ReutersThomson Reuters held a discussion last night (Nov 12, 2009) in their Hong Kong headquarters on the finer points of High Frequency Trading (HFT). The panel was moderated by Mike Powell the Global Head of Enterprise Information at Thomson Reuters. The panelists invited to speak represented a large swathe of the electronic trading vertical including buy side, sell side and technology. The discussion lasted about an hour and was well attended by an equally diverse cross-section of the industry.

It was generally accorded that high frequency trading was not new a practice but has been under the media spotlight recently garnering attention.

High frequency trading was defined as a data intensive, latency sensitive, high volume, low margin activity where holding periods were very short and positions were rarely held over night. High frequency traders were defined more by the services demanded than by how quickly orders were placed.

High frequency traders are made up of international banks, large hedge funds and smaller boutiques each facing a high cost of entry. Tick data is expensive, slippage matters and investing in sophisticated technology presented a barrier as well. It was raised that human capital investment was lower and greater emphasis was placed on IT. More money was being traded with less people.

There was some divergence on the opinion that large firms tend to be less restricted to trade anywhere where small players were more agile and free to enter any market.

Volatility and liquidity were the prime drivers of High Frequency trading. Also, the “Market landscape” played an important role in whether HF trading was a viable pursuit. For example, it costs 30bps to sell in Korea making this kind of trading prohibitive when the battle is being fought for the spread. India and Taiwan do no allow day trading.

The ability to process the large flows in the post-trade was another factor contributing to the increase HF trading.

An interesting point was raised about cultural influence as a driver of High Frequency trading. Japan, for example, has a more conservative fundamental approach to the investment process. Trading based on quantitative decision making is unusual. Also a “Quant Culture” supports the notion of HFT driving trading innovations and ideas.

Co-location services, it was agreed, was on the rise in Asia as well both at the exchange and broker level. The Australian Securities Exchange has been providing co-location services for some time and the Tokyo Stock Exchange just announced and expansion to its facilities.

There was very little said about risk other than that sophistication is increasing.

All and all it was a captivating discussion providing food for thought on a growing and exciting segment of the electronic trading industry in Asia.

About the Author


Displaying 6 Comments
Have Your Say
  1. Paddy Osborn says:

    Looks like an interesting discussion. While the vast majority of HF trading is done off market price data, does anyone out there see increasing opportunities in HF trading off macro-economic news? e.g. Low latency Algo trading off payroll data, CPI, PPI, housing data, etc.

  2. Steve says:

    Hi Paddy,

    Thanks for the feedback. To be honest I am not sure what the demand from the buy side is on this input to algo decision making. I have come across a few vendors Reuters and MNI (Eurex company) putting up their shingle as it were offering this feature in Asia however. I am attending Tradetech in Sinapore this week and will certainly be asking around about it. Great blog article. Hope you singed up for the Weekly Bulletin.

  3. Paddy says:

    I’d be interested to hear your feedback from the Singapore event. From what I hear, trading off macro data is being done already by most major banks and hedge funds. Mostly in US for now, but looks to be spreading East through Europe into Asia. How do I sign up for the Weekly Bulletin? Thanks.

  4. Steve says:

    At this link

    I see you figured it out. Thanks for the feedback and look for more information at

  5. Tom says:

    Two things Steve:

    1) The term “low latency” usually implies high frequency. But one of our clients pointed out this distinction: He has our low latency software, but isn’t high frequency, trading only once or twice a day. He calls his trading “quick response”, and that’s why he needs low latency data.

    2) Macro Event affects are relatively long in duration, thus diluting the need for micro-second response times. We don’t see much interest in this area.

    Flume Data

  6. Steve says:

    Hi Tom,

    Thanks for the note. Interesting low frequency low latency requirement. I guess its up to the sell side to provide all types of services to all types clients. Anyone else have a similar LFLL scenario?

Leave a comment

XHTML: You can use these html tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>

IRP Journal

IRP Journal



Poll results are published in our Weekly Newsletter -->
All Rights Reserved WIld Wild Web Limited