TECH FIRM PROMISES LOW-COST, FLEXIBLE ECONOMIC SCENARIO GENERATOR
UK-based financial analytics provider Financial Canvas is promising insurers a flexible and easy-to-use economic scenario generator (ESG) that could save them hundreds of thousands of pounds a year.
Established by Chris Squirrell, a former head of ALM modelling and stochastic valuation at Aon Consulting, Financial Canvas has been working with pension schemes, life insurers and advisors since 2013 to provide analytics solutions, including an alternative to the off-the-shelf ESGs that currently dominate the market.
Many of the existing ESG solutions are difficult to understand, slow to calibrate and are hard to use for decision-making, Squirrell said. “We wanted to create a model that had all the features that people wanted, but is easy to set assumptions.
Squirrell said the firm’s ethos is to put users in control of the analytics.
“That’s done by making models very visual – you can drag-and-drop components, plug them together and build models. We will do that with clients, using our toolkits, and build a model they can understand.”
In practical use, modelling teams often struggle to their make ESGs conform to their assumptions, and recalibration can be a lengthy process that involves tweaking esoteric factors that end up influencing other elements.
By providing the ability for rapid recalibration, Financial Canvas users can test the robustness of their decisions to the assumptions in the model. “Traditional ESGs can be out of date before the calculations are complete. You cannot apply the existing economic situation to an old model,” Squirrell said.
Visualisations are also important for the outputs from the models, he said. “The analytics may be relatively easy, but what is really hard is being able to use them to make decisions and add value. We put a lot of thought into interactive dashboards with a clear narrative to help with this.
“In the context of ESG this might mean, for example, how does changing an assumption affect our required capital? It also means providing people with engaging tools to understand the nature of the distributions they have generated.”
The software is based on Matlab, allowing users access to a well-supported library of pre-built mathematical models and optimisation algorithms.
Financial Canvas currently offers a real-world implementation and is set to release a risk-neutral model.
The product has a base platform fee and three levels of licence: a model building licence; an administrator licence, allowing oversight of the models; and a dashboard-user licence.
“None of our clients pay more than £100,000 ($130,000) a year for everything,” Squirrell said. “We are in the business of gaining clients and our overheads are lower than some of the big firms.”
The firm also offers consultancy to help users alter the models.
“In practice, for a new client, we would be asking what they wanted and we would deliver that to them for a modest fee,”
he added.
Christopher Cundy