Today we have spoken to Fredrik Davéus, the founder and the CEO of Kidbrooke, about his experience of leading digital transformation in a number of prominent financial institutions in the Nordics. Fredrik walked us through the initial requirements to meet before starting a project, the components of financial analytics to look out for and how a standardised implementation process might look like.
Where to begin and which initial technical capabilities to consider as a requirement when adding digital financial analytics?
The type of analytics you need is typically connected to your customers' needs and, consequently, of your own team. Therefore, it is useful to start by taking a step back, analysing your customers and identifying the information and types of decision-support they might benefit from.
As a next step, consider the following points:
- Would you like to project the entire balance sheet of the customer or only specific parts of it?
- Is it important for the analytical tool to consider your own house views in addition to more general historical characteristics and consensus views?
- How granular and realistic would you like the underlying asset modelling to be?
- What is your technology stack like? Do you have a platform or application into which a new analytics API can be integrated, or do you also have to find and implement such a platform itself?
- How many customers do you need to serve? Self-service solutions may require a more performant and scalable set-up compared to advisory tools used in physical meetings.
- How are you going to manage compliance around the decision support you provide to your customers? The more sophisticated the analytics, the higher the demand for transparency, ability to audit and validate the underlying algorithms.
What are the main components to consider of a digital financial analytics module?
It might be useful to think of the analytical technology in terms of three main building blocks:
Scenario generation – an element driving all assumptions about future developments of financial markets. Scenario generators range from being very simplistic, e.g. only considering expected returns for each asset class, to more sophisticated, driven by a multi-period stochastic factor-based model capable of reproducing all well-known characteristics of financial returns.
Balance sheet evaluation – a layer translating the fundamental assumptions from the scenario generator to customer-specific projections of their portfolios or balance sheets into the future. If needed, holistic support is an important aspect to look out for in this area.
Decision-support tools – a module producing the proposals and decision-support material constituting the core of the value-add you provide to your customer. If you have a discretionary business model, you might not need this component at all. In contrast, if you are building any self-service proposition that would automate the decision-making, you need a structured way of allowing the “machine” to generate such output.
Do you require a dedicated development team to implement digital capabilities?
The answer depends on the size and the level of ambition of the firm. For example, Kidbrooke has large customers, such as Skandia Life, which have dedicated teams working on developing various aspects of their technology stack, and smaller ones like Evida, which has chosen to outsource all the required development work to their partners.
How flexible are the APIs in the context of various infrastructural set-ups? Is it possible to integrate them with existing systems?
OutRank is a standard RESTful API that comes with OpenAPI specifications; thus, it can be integrated almost everywhere, e.g. directly into the user interface layer or into backend applications orchestrating the customer or business processes.
Do you have a standardised process of setting up the software, and what stages does it involve?
The standardised process typically involves three stages:
Extensions of functionality (optional) – customer-specific functions or customisations (in terms of country-specific tax logic, for instance) may have to be added to our software.
Configuration – the alignment of assumptions ranging from outlooks or house views affecting the scenarios to mapping specific investment products and fee structures, etc.
Integration – creating the API- calls from the application on the customer side to the OutRank API. OutRank comes with modern OpenAPI specifications that allow you to generate client software development kits in your preferred language. If the customer journey is clearly defined, integration is often very quick. There are examples of integrations that have been done in less than one week.
In addition to the above, our customers sometimes involve Kidbrooke in developing their desired use cases and customer journeys. In such cases, the process becomes less standardised and depends on the project's size, the number of stakeholders and the decision-making culture at the customer’s side.