Introduction to the Risk Matrix and Algorithms

Data is everywhere. It's where you can see it, where you can hear it, feel it, taste it and smell it.

When we are building a financial crime risk matrix, we use all of those senses to collect data.

But there is also data we cannot see, hear, feel, taste or smell. We look for that data by extrapolating from what we have. And sometimes there is data that we cannot identify by extrapolation.

People laughed at Donald Rumsfeld for his "unknown unknowns" speech but we use it every day when we are assessing risk.

In this workshop, working as a team, we will build a risk matrix for a fictional regulated business. Why? Because then you will understand so much more about financial crime risk RegTech.

What are the challenges?

First to identify relevant data sources - and yes, the physical world provides us with a steady flow of data. Everything from the shoes someone wears, the bag he carries, the perfume she wears all combine to give a powerful impression of a customer.

Exercise

Exercise - write down data

Take five minutes to write down everything you notice in the image of a girl on a train. Include your impressions.

Put your note away and don't look at it. In two days, come back to this page and do the same exercise again. Then take out your first note.

Compare your two notes. Are they the same?

Now write a third note: describing to yourself what those differences were and why you think they arose.

Finally, recognise that your data source was constant. Imagine if the data sources had been similar but not identical.

All results of computations, be they human or computerised, depend on what data is available, the quality of that data and the weight that data has in relation to the whole dataset. Such computations assume that there is nothing unknown, or unknowable, that would affect the results.

A risk matrix records all available data, it affords a weight or value to each piece of data and it builds a risk rating which is measured against the company's risk appetite.

And that risk appetite is a can of worms that must be unravelled if the model we build is to have any real-world value.

The second challenge is to recognise what technology is capable of doing with the data it has.

We will use data, in algorithms we will build, to analyse data and to ascertain the quality and nature of the reports generated.

There's no big secret: we will show that machines are brilliant at assessing the fact of compliance - does the company do this/does it not. We will also show that algorithmic analysis can never do better than provide a statistical probability that the data it has generates a score that says "look at this."

Don't have apoplexy: I'm talking about algorithms, statistics, probability like they are as common as muck. And guess what, they are. There is absolutely no reason to be intimidated by them.

You've used algorithms from the moment you were born - it's just that no one called them that.

So we'll demystify them, build some and equip you to understand the technology : it all uses exactly what you will learn in this course: the only difference is to operate at speed and scale.

Who should attend - Introduction to the Risk Matrix and Algorithms

Practitioners in financial crime risk assessment and management.

Includes MLROs etc.

Internal Audit

Their staff at all levels.

Which industries should attend Introduction to the Risk Matrix and Algorithms

All regulated industries.

All commercial concerns in trade.

Regulators

Enquire

For more information and the course outline.

Your email will be validated against MX Records.

Corporate email address only.

Click here

back to

Developed & Designed by Alaa Haddad