Investment manager accelerates development of quant strategies with one data platform2017-03-14T11:45:45+00:00

Project Description

Investment manager accelerates development of quant strategies with one data platform

The asset management arm of one of the Top 10 Swiss Banks uses GAIN Quant DB to industrialize the data preparation activities of its “quants” research team, by automating the aggregation of index and market data from multiple sources (using the Thomson Reuters QA offering that includes DataStream, IBES, Worldscope as well as MSCI indices) to feed MatLab, the portfolio modelling application.

Quantitative Analysts – also called “quants” – design and implement complex models to create new investment strategies. The process involves the capture and preparation of large volumes of time-series from various sources to research and back-test. The goal is to validate the models. The data preparation is a crucial step to guarantee the best possible accuracy of the data underlying of new quantitative models before new funds are launched.

In many Asset Management companies , each quant is developing its own tools, scripts, databases, spreadsheets, and  this data preparation work has historically been manually intensive, time-consuming, error-prone and not fully documented.

In worst cases, the know-how on the datasets and tools leaves upon individual’s departure, leading to a high operational risk for the institution. Successful Quants are valuable and can be headhunted by the competition – together with the know-how, the tools and the data they have been using to develop their models on their desks. Institutional investors increasingly demand for more professionalism in quantitative investment, still perceived as a lucrative but secretive area of asset management.

The client needed to automate the data preparation activities. “We needed a powerful solution enabling us to easily aggregate index and market data from multiple sources including Thomson Reuters QA Direct as well as MSCI and Bloomberg to feed MatLab. AIM Software’s GAIN Quant DB was the perfect match” says the Head of Investment Management of the firm.

“We needed a powerful solution enabling us to easily aggregate index and market data from multiple sources including Thomson Reuters QA Direct , MSCI and Bloomberg to feed MatLab.”

Head of Investment Management of Swiss Top Asset Management firm

The client uses GAIN Quant DB to automate the once tedious task of data preparation and as a result the customer gains in efficiency and model accuracy and keeps control of the know-how about the models developed by their “quants”.

GAIN Quant DB acts as a central Research database that automatically collects and cross-references data from various sources, thus eliminating the risk of error-prone manual consolidation. Snapshots of market data are created several times a day using a vendor-agnostic approach, thus reducing the risk of “look-ahead-bias” – a known issue in commercial databases happening when historical facts such as returns are revised under the light of new facts not known at the initial time of writing. The snapshots are stored centrally and made available to all the quants for the development and validation of their models.

As a result, the time series available to the quants are more accurate and can be traced back; at the same time the central database requires ten times less storage. The hours saved per day allow research phases to be reduced by one to three months. Even more important: The validated models can be launched within days instead of several weeks.

“Quants are faster in getting the data they need and can focus on what really matters – creating better investment strategies. This reduces our operational risk and our reliance on individual know-how, while highly contributing to our mission of stable and efficient investments for our clients”, comments the firm’s Head of Investment Management.

Investment manager accelerates development of quant strategies with one data platform

The asset management arm of one of the Top 10 Swiss Banks uses GAIN Quant DB to industrialize the data preparation activities of its “quants” research team, by automating the aggregation of index and market data from multiple sources (using the Thomson Reuters QA offering that includes DataStream, IBES, Worldscope as well as MSCI indices) to feed MatLab, the portfolio modelling application.

Quantitative Analysts – also called “quants” – design and implement complex models to create new investment strategies. The process involves the capture and preparation of large volumes of time-series from various sources to research and back-test. The goal is to validate the models. The data preparation is a crucial step to guarantee the best possible accuracy of the data underlying of new quantitative models before new funds are launched.

In many Asset Management companies , each quant is developing its own tools, scripts, databases, spreadsheets, and  this data preparation work has historically been manually intensive, time-consuming, error-prone and not fully documented.

In worst cases, the know-how on the datasets and tools leaves upon individual’s departure, leading to a high operational risk for the institution. Successful Quants are valuable and can be headhunted by the competition – together with the know-how, the tools and the data they have been using to develop their models on their desks. Institutional investors increasingly demand for more professionalism in quantitative investment, still perceived as a lucrative but secretive area of asset management.

The client needed to automate the data preparation activities. “We needed a powerful solution enabling us to easily aggregate index and market data from multiple sources including Thomson Reuters QA Direct as well as MSCI and Bloomberg to feed MatLab. AIM Software’s GAIN Quant DB was the perfect match” says the Head of Investment Management of the firm.

“We needed a powerful solution enabling us to easily aggregate index and market data from multiple sources including Thomson Reuters QA Direct , MSCI and Bloomberg to feed MatLab.”

Head of Investment Management of Swiss Top Asset Management firm

The client uses GAIN Quant DB to automate the once tedious task of data preparation and as a result the customer gains in efficiency and model accuracy and keeps control of the know-how about the models developed by their “quants”.

GAIN Quant DB acts as a central Research database that automatically collects and cross-references data from various sources, thus eliminating the risk of error-prone manual consolidation. Snapshots of market data are created several times a day using a vendor-agnostic approach, thus reducing the risk of “look-ahead-bias” – a known issue in commercial databases happening when historical facts such as returns are revised under the light of new facts not known at the initial time of writing. The snapshots are stored centrally and made available to all the quants for the development and validation of their models.

As a result, the time series available to the quants are more accurate and can be traced back; at the same time the central database requires ten times less storage. The hours saved per day allow research phases to be reduced by one to three months. Even more important: The validated models can be launched within days instead of several weeks.

“Quants are faster in getting the data they need and can focus on what really matters – creating better investment strategies. This reduces our operational risk and our reliance on individual know-how, while highly contributing to our mission of stable and efficient investments for our clients”, comments the firm’s Head of Investment Management.

Related Client Stories

Ask About Our Solutions


CONTACT US

Send us a message and one of our team will get back to you.

REQUEST A DEMO

See our products in action. One of our specialists will contact you.

 


SUPPORT

Are you in need of support? Our dedicated team is here to assist.

 


LOCATIONS

Click here to see our office locations.