RoboR8.com, based in Singapore, is led by Simon Christofides, who formerly headed credit index trading at Goldman Sachs in London and served as chief investment officer of SW7 Asset Management in Singapore. He sees an opportunity to open credit scoring to a wider community — professional and non-professional alike — and provide continuous monitoring of company ratings in a centralised database.

Available now as a free version, RoboR8 (pronounced robo-rate) is a company credit scoring system that uses a proprietary machine learning algorithm called Migo (machine intelligence via graph optimisation).

“The output of RoboR8 is congruent to the classification used by the major rating agencies,” says Christofides. “Migo combines a classification tree approach, which can be proven mathematically to be optimal, with graph-theoretic algorithms for factor selection and evolutionary systems for forest aggregation. The result is an engine that we believe to be superior to and more robust than existing automated scoring systems.”

At launch the system will include models for 8,000 US, European and Chinese companies, spanning around half a million ratings that the system updates continuously. Screens show a distribution of appropriate ratings after layering company financial statements with macroeconomic inputs chosen for each industry sector.

A paid-for service providing full risk analysis will be available in due course.