Author: Jonathan Phillipo, Liqueo Senior Consultant
Data Scientists have been an integral part of the investment process within asset management for many years. Usually, Quantitative strategies, those which use mathematical models to make investment decisions, tend to make up a large percentage of all investment funds. However, quants have typically been involved at the start of the investment process rather than at execution. Now, more than ever, Buy-Side Dealing Desks are turning to Developers with programming and analytical skills to help navigate and interpret the huge amounts of data at their fingertips.
Having a dedicated Data resource on a Dealing Desk who can assist with ad-hoc analysis of Transaction Cost Analysis, Liquidity impact, and Pre-Trade insights offers an edge over other Asset Managers.
The main programming languages used for Data Science are Python and R. R is a programming language built for statistical analysis and is popular with academics. Python has become the de-facto coding language of finance. The latter has many libraries for statistical analysis and the syntax borders on natural language - making the barrier to entry lower than other programming languages. Both languages boast readily accessible open-source packages able to handle the needs of finance.
The necessary companion to programming languages, a database, is an important tool in the data mining pipeline. Popular open-source options include SQLite, Postgres, MySQL or MongoDB for unstructured data. In terms of commercial offerings, MS SQL Server has come a long way in recent years and is now a very strong product for structured data storage. Depending on the size of the data sets, Big Data databases such as Hadoop Distributed File System (HDFS) with Spark or Snowflake can be used to query the data as needed.
In addition to technical skills, any Developer being placed on a Dealing Desk would ideally have a broad range of soft skills, with communication and presentation skills being key requirements. The Developer should have a thorough understanding of the business and technical domains and be capable of presenting to and collaborating with lots of different people both within the organisation and externally.
Sell-side counterparts more frequently ask for a detailed breakdown and analysis of the data presented to them. SThis typically happens on a monthly or quarterly basis in the form of a broker review meeting, where sell-side counterparts are looking to understand where they rank in terms of the number of trades and value of trades to the asset manager, and how well they have executed the trades.
Transaction Cost Analysis (TCA) data, showing various metrics such as Volume Weight Average Price (VWAP) and Implementation Shortfall, require formal definitions so all parties have a common understanding of the analysis. This detailed data mining is time-consuming and requires specific skills combining programming, statistics, and data visualisations. Having dedicated resources who can extract these insights and have conversations directly with sell-side representatives is invaluable.
Being able to look at historical data and use machine learning to provide insights around liquidity gives traders a pre-execution edge over peers. The ability to flag trades as having a larger liquidity impact allows traders to split trades into high and low touch orders, with a view to automating the easier trades. This gives focus to the orders for which traders can add value, by sourcing liquidity, getting trades done and obtaining the best price.
Many modern Order Management Systems include the ability to automate certain aspects of execution. For example, Charles River, a popular Buy-Side Front to Back solution, has functionality for Auto Routing of orders based on rules. These orders can be sent directly to Algos, or Financial Information Exchange (FIX) Venues for auto execution. This automation frees up the Buy-Side trader to add value to the more difficult trades.
This will be an interesting space to watch and could see the role of buy-side dealers change even further than it has in the last 20 years. Are data engineering skills becoming foundational skills? Might the Developer on the Dealing Desk also perform some execution duties as well?
Here at Liqueo we specialise in both Front Office and Data and have dedicated resources that can help define operating models of the future. Contact us for more information.
We use necessary cookies to make our site work. We'd also like to set analytics cookies that help us make improvements by measuring how you use the site. These will be set only if you accept.
For more detailed information about the cookies we use, see our Cookies page.
Necessary cookies enable core functionality such as security, network management, and accessibility. You may disable these by changing your browser settings, but this may affect how the website functions.