About Us

Our Mission

We are a RegTech start-up with big ambitions. We believe in better. Good decisions need insightful analysis and this only happens when you know where and how to look for and within the data. Our mission is to help our clients identify emerging risks and either avoid losses or improve performance. We do this by providing a unique and accessible risk scoring data service that will assist in the early identification of entities at risk of failure due to accounting manipulation or fraud.

Our Values

At the heart of what we do, we are a data science company. We love data. We value its unbiased nature, neither slave nor master to opinions. We value meaningful insights that can be derived from good clean data and we connect the dots in between to help you make those insights powerful. Above all, we value transparency and strive to deliver a unique service in a manner that demonstrates this value.

Integrity lies at the core of our values. Too frequently we have seen all manner of accounting malfeasance by companies with serious consequences for stakeholders, customers, suppliers, employees, etc. We designed our system to identify these actions very early on. It is therefore fundamental to our business that we too ascribe to the highest standards of honesty and inegrity.

What Is The MRA?

Manipulation Risk Analyser

The Transparently Manipulation Risk Analysis (“MRA”) provides a score from least likely to most likely for companies that exhibit “red flag” risk factors for possible financial manipulation, fraud or mismanagement.

Taking learnings from existing techniques, such as forensic accounting, automated signal detection approaches for financial crime and machine learning, we hunt for the digital DNA likely present in the manipulation of accounts and financial crime.

These signals are often deeply buried in vast quantities of diverse data and hard to detect by the human effort of forensic accountants, auditors, stocks analysts, credit analysts or even experienced investors. The power of machine learning algorithms makes this behaviour much more transparent.

Large datasets are employed to identify leading risk signals for a wide variety of ”red flag” characteristics, which in turn are employed to identify clusters of risks and ultimately an overall risk score. This manipulation risk score represents the joint probability of manipulation and corporate failure resulting from that.

How Does It Work?

The boring bit

Providing detailed risk assessment and analytics for over 22,000 publicly listed companies in the US, China, Hong Kong and Singapore, requires hundreds of millions of data points. More markets will be added in the coming months.

The system generates an overall percentage risk (0-100%), representing the probability of corporate failure due to accounting manipulation and fraud.

The MRA is hosted on Google Cloud and meets industry best practices for security.

The geeky bit

We use the latest data science techniques in machine learning, and forensic accounting processes, to solve multi-dimensional problems at this scale of complexity. We built an array of financial models designed to replicate the investigative activities of forensic accountants, activist short sellers and auditors.

The system has been extensively evaluated for its accuracy and robustness over multiple decades, across different markets, independent datasets and myriad other ways. Consequently the system produces highly accurate results with a long lead time to market awareness of problems (on average, 2-3 years).

The clever bit

Most importantly, the MRA is not just a “black box”. The challenge with most machine learning is that users only get results and not explanations. The MRA provides users information on what to do next, what to investigate, what questions to ask and identifies where in the accounts things don’t look quite right.

Our system automatically generates results in a customised report designed to not only signal failure risk from accounting manipulation and fraud, but also explain where and how a company is manipulating its books.