The difficulties early stage hedge funds face in gaining a foothold through access to exchange data and other venues can be partially offset by the rise of ‘data science as a service’ solutions, according to a study from BMLL. The firm’s chief product officer, Elliot Banks, spoke exclusively to BEST EXECUTION on how BMLL helps early stage and start-up hedge funds focus on research and finding an angle to generate alpha, rather than devoting resources on intensive in-house processes associated with curating and cleansing data.
Early stage hedge funds’ need for substantial data engineering infrastructure, which does the heavy lifting of research and computational and continuous model calibrations, can be be met by data science infrastructure as a service, which brings down barriers to entry for such firms looking to leverage critical data engineering infrastructure at speed and reduce their time to market.
Banks told BEST EXECUTION: “Historically, hedge funds have done it all in house: they have built the data science capabilities themselves. However, we have seen that, over a period of time, many parts of this process are outsourced, such as the processes related to trading, infrastructure, data and historical data, as well as the research component.”
“Young funds, such as start-up funds, want to focus on the value-add. They realise that they do not need the burden of the in-house processes associated with curating and cleansing data, or managing and storing multiple copies of historical market data. If their core competency is research and finding an angle to generate alpha, the effort needed for data curation and data engineering is not where they want to spend their time and resources,” Banks added.
“What we offer early stage and start-up hedge funds is historical data packaged with an analytics library and the ability to query and analyse the data. They can put that into their system and run computations every single day. They can use APIs, scale their compute power or use the cloud to run their own research when they need it. We give them both the data and the environment to support their trading strategies – that is data science as a service,” Banks concluded.
Additionally, data licensing fees and data security costs can be prohibitive, particularly if the fund relies on traditional market data directly from exchanges and other trading venues. A new case study from BMLL outlines how as the costs and expertise needed to conduct effective research and backtesting grow, in-house infrastructure may struggle to keep up with ballooning data volumes.
So-called out-of-the-box ‘Data Science as a Service’ solutions allow quants to gain access to a full depth order book, including Level 3 data, freeing up their resources to focus on strategy and research at the hedge fund.
By using a pre-built cloud-based platform from a provider, early-stage hedge funds can gain access to engineered, high-quality, granular market tick data in a secure dedicated sandbox environment for carrying out research, signal detection and back testing, without the need to set up direct data licences themselves, nor incur additional licensing fees to access the full depth venue order book (Level 3) data.
© Markets Media Europe 2023
