The webcast, Latency and Liquidity, is sponsored by IBM and Finextra and supported by the FISD is now available on demand.
Our panel of industry experts includes:
- Elizabeth Lumley, special projects editor, Finextra (moderator)
- Keith Bear, director of business development for Financial Markets, IBM
- Jack Vensel, MD, head of electronic Trading, EMEA, Citi
- Kee-Meng Tan, MD, Knight
- Gregory Smith, vice chairman of Chi-X Global
There is no question that most banks take reduction in latency very seriously. Add in factors such as direct market access, co-location and the race to find an edge in a sea of fragmented liquidity; many banks are drowning in market data, networks, algos and data centres. With the current global recession new regulations covering high frequency traders and dark pools hang like a shadow from the US over most of Europe.
Regulators and technology is a touchy subject. Much has been made around whether regulators have the means and the resources to monitor markets that rely on ever changing and sophisticated technology.
The webcast debate looked at a number of issues surrounding Speed, Fragmentation, Complexity and Regulations that affect electronic trading.
Choice quotes from the webcast:
Jack Vensel, Citi
"Keep in mind that 100% of order flow from people who do not have a SOR goes to the primary exchange."
Kee-Meng Tan, Knight
"Let's not kid ourselves, the exchanges are not making this investment [in data centres] to help make trading faster, the truth of the matter is they are basically investing in these facilities to make money, because their trading revenues are being eroded and they are under pressure - it is just another source of revenue."
Keith Bear, IBM
"There is a level of diminishing returns as far as latency is concerned, there comes a point where the last micro-second is going to be that much more expensive to achieve and if its only part of the puzzle overall, then that also has to be taken into context, eg in adding more capacity."
Gregory Smith, Chi-X
"80% of execution strategies are not really that dependaent on the current state of latency; , what they care about is getting best execution - there are a lot of dimensions to it; , it is not just about latency."