Big Data Sharpens FX TCA at ITG
Trading analytics for foreign exchange are somewhat sparse historically, but ITG is using new data sources to hone its analytics.
“We’re improving the tool that we provide to our clients by analyzing the wealth of available tradeable quotation data, which wasn’t available to analysts four or five years ago,” said Jim Cochrane, director and senior product manager at ITG.
Before the largest FX banks adopted an agency execution model, FX traders accepted risk-transfer prices or a fairly significant T-Costs.
But now that the traders are taking on more currency risk themselves, they are examining every variable that makes up their transactions costs, according to Cochrane.
The brokerage is using ‘big data’ analytics to understand the available market data currently available.
“I think the study that we’re doing in the FX market is really the challenge, because of it is difficulty of getting representative data sets and universal data sets,” he said.
Two areas which ITG is investigating is the development of a benchmark rate that will help FX traders find a more accurate price when calculating implementation shortfall. The second project is determining when and where are the best times and locations to execute implementation shortfall.
In order to calculate its benchmark, the brokerage is including pricing data from FTSE Cürex as well as from the Lmax Exchange’s and Nasdaq’s market data feeds into its indicative price calculations.
Those prices might be higher than the riskier last-look quotes, but they have the risk-transfer price built in to them.
By using this new benchmark, ITG can start examining the results of FX trade executions using peer-level data to determine more accurate transaction costs.
Predata uses alternative data and machine learning.
VeloCity will accelerate fintech tailored to asset management.
Combining human and machine intelligence produced the best results.
Dutch fund manager goes beyond traditional economic numbers.
First wave was finding data; second wave will be about interpreting data.