Client
A leading private wealth manager based in California with $8 billion in assets under management and a client base of high net worth individuals and families. The Client offers highly-customised wealth management solutions across all asset classes (public and private) with a focus on integrating complex multi-generational investment considerations.
Problem
The investment team’s proprietary asset allocation and portfolio optimisation model and processes were entirely built, run, and maintained in Matlab by a single third-party developer. This stifled their development cycle and made model changes extremely inefficient and time consuming.
The creation of new charts/visualisations for clients would take months in most cases and the code underlying these analytics was not accessible in-house, introducing key man risk. The Client’s optimisation tool took up to 3 minutes to render output.
The Client hadn’t previously developed a risk model, which limited their ability to perform forward-looking analysis. They couldn’t dynamically demonstrate how changes to current allocations would impact risk, return and liquidity, ultimately impacting client objectives. Performing bespoke/what-if analytics that reflected their investment team’s philosophy was not possible during client meetings.
Additionally, the Client was unable to monitor client portfolios at scale. The proprietary system was not programmatically connected to their book of record vendor, leading to heavy reliance on spreadsheets and fostering mistrust in the accuracy of client portfolio data used in calculations.
Solution
Jacobi’s investment engineers conducted a thorough, line-by-line review of the Client’s existing
MatLab code. This code was then re-written in Python, effectively rebuilding their proprietary
Optimiser in a modern, scalable format. A detailed parallel study of output was then performed by Jacobi investment engineers to ensure the new calculations matched the previous third-party maintained system.
The Client could leverage Jacobi’s “out of the box” risk model to immediately address their inability for dynamic client engagement, incorporating “what-if” scenarios that could be used in real time by investment and client-facing teams. This eliminated hours of customised work to address simple client questions, previously performed between book of record data exports, Excel and Powerpoint.
Jacobi worked with the Client’s book of record vendor to establish a programmatic feed of their client portfolio data into the platform, eliminating a previously manual upload process. This would allow the Client to have a fully functional, white-labelled platform which incorporated their proprietary optimisations and Jacobi’s risk model - all integrated in an easy to use user interface.
Most importantly, the Client would no longer be reliant on a third-party to change their models and could roll-out the platform to client service teams across their organisation.
Outcome
The Client benefited from improved process efficiencies which allowed them to focus on their core objectives: delivering outperformance, exceptional client service and growing their business. The optimisation calculation which previously took 3 minutes is now rendered in under 60 seconds.
Workflows that were captured in disparate processes (e.g. optimisation, ex ante scenario generation and stress testing, ex post portfolio analysis) are now integrated in one platform and the time devoted to reacting to client queries has reduced. The Client now has full visibility across all their client portfolios, with the ability to do ex post and ex ante analysis, plus integrate this with their client reporting and engagement framework.
About Jacobi:
Jacobi's technology has its roots in institutional investment management and brings together investment expertise and a market-leading technology platform. Headquartered in San Francisco with offices in London, UK, and Brisbane, Australia, the company is led by a team of experienced investment professionals and engineers.
