Financial firm TD Securities Automated Trading (TDSAT) is using Google Cloud Platform for trading fixed-income bonds.
In addition, parent company Toronto-Dominion Bank has signed a multi-year deal with the cloud provider.
In a recent blog post by managing director and head of automated training, Matthew Schrager, the subsidiary of Canadian banking group TD Securities revealed how it has patterned with Google to use technology, automation, and quantitative modeling for trading in the municipal bond market.
In the post, Schrager said: "Google Cloud’s offering has enabled us to construct a research platform that is one-of-a-kind in this space, with enough horsepower to drive the massive research workloads associated with our data-heavy approach," adding that the scalability of the cloud compared to on-prem servers has also enabled the company to only pay for high-performance infrastructure when it is needed.
Historically, bond markets have been complicated in ways that prevent automation. For example, a company generally has one associated stock, but can have hundreds of unique bonds. In the municipal bond market, there are millions of unique securities.
Individual bonds also have less trading information, meaning datasets are more fragmented and sparse, and trades occur only once every few weeks.
The blog post explains that fixed-income strategies "involve learning over all instruments simultaneously. This reduces the degree to which the problem can be parallelized by several orders of magnitude, which has significant implications for algorithm selection, compute infrastructure, data organization, and more.
To get around this, TDSAT uses Google Cloud infrastructure and kubernetes which can drive hundreds of thousands of concurrent research jobs at all hours.
According to TDSAT, the main challenges included the need to run dozens of experiments per day, each of which is compute-intensive, requiring thousands of concurrent cores, and needing fast access to large datasets. Finally, the workload is bursty and unpredictable.
"An on-prem compute farm of sufficient scale would require an initial outlay of tens of millions of dollars, plus hundreds of thousands of dollars in annual support. It would also take time to build – we’d have to find a data center to house the servers, procure the hardware, hire ops staff, wire it all up, etc.," Schrager said.
By using Google Cloud, TDSAT has been able to provide competitive pricing on hundreds of thousands of fixed-income securities, and is now one of the most "algorithmic bond trading groups in the world."
The wider company has also inked a 'multi-year contract' that will enable it to "respond quickly to changing customer expectations by rolling out new features, updates, or entirely new financial products at an accelerated pace.”
