About Us
We are APDS Analytics, a boutique retail credit risk analytics & modelling consultancy set up in 2016 as a vehicle to offer state of the art analytics to the global financial services industry, through contracting and freelance projects.
​
Based in London, we have worked with major UK high street banks (mortgage product launch & IFRS9 modelling), a Middle Eastern bank that was aiming to lift their share of wallet, a Thai sub-prime lender controlling origination risk, a Mongolian start-up digital bank and a top 4 NZ bank (analytical marketing comms) or a CRA that wanted to introduce Alternative Data Scoring for the unbanked population.
​
As an independent firm we can also offer model validation services, to help you comply with the ever increasing model risk guidelines.
​​
We are currently delivering cutting edge analytics, utilising traditional and alternative datasources, to a number of banks in Cambodia with the overall aim of boosting Financial Inclusion and Economic Growth.
​
Additionally, we offer model development training and knowledge transfer sessions around Scoring Concepts, Model Monitoring and Validation.
The company is led by Matthew Freeman, a 30 year veteran of the banking scoring / analytics sector, with experience gained globally.
Meet Us
Email Us
Credit Scoring & Propensity Modelling
(Empirical Data Driven Models)
**** Click on the pictures to go to that area *****
Open Banking Analytics
(Risk, Collections & Marketing)
Generic / Expert Scorecards
No Data Models and Roadmap to Empirical
Monitoring, Validation & Model Risk
Retail Risk Management Framework
Innovation & Financial Inclusion
Bureau / CRA Analytics
Training
Partners
Papers
Global Experience Applied Locally
Across the globe our consultants have delivered varied projects, adding experience and knowledge that can be applied to new projects, whatever the challenges
Partners
coming soon
including
Alper Consulting
Pro Data Science
Analytrix Engine
CrePass
and Others
Past and Present Clients
Since 2016 we have worked with a number of organisations, directly or through partners, for which we are very grateful.
​
Let's add to this list, click on the link below
Analytically Powered Credit Decisioning
Origination Scoring
​Determining from a risk perspective which (new) customers to accept under what terms
(assessing how rejected customers would have performed had they been accepted)
Behavioural Scoring
Behavioural Scoring
Determine how to treat existing customers from a extended lending perspective, early collections treatment etc.
Collections Scoring
Collections Scoring & Payment Projection (Recovery Modelling / Pseudo LGD)
determining the treatment of customers that have fallen into delinquency and how they roll between delinquency buckets, with the aim of maximising the amount of debt collected or recovered, at the same time as prioritising the use of organisational resources.
Colls Scoring, click to download
Benefits of Scoring
-
automated and quick decisoning
-
consistent decisions made
-
ability to control the flow of applications into fulfilment or the flow of accounts into collections
-
improved customer service as the bank proactively handles customer decisions based upon behavioural traits
-
utilising summarised customer data allows the bank to determine future customer needs and offer appropriate products and services at the right time
-
Maximises the amount collected / recovered from defaulted accounts
-
Operational scoring will also provide a platform for the bank's regulatory models, potentially lowering capital (IRB) and provisions (IFRS9)
-
The development & use of both application and behavioural scoring engenders a deep portfolio understanding across the portfolio teams
-
APDS Consultants will work with your risk analytics team to deliver scoring solution whilst ensuring awesome knowledge transfer
​
Email Us
Take me home
Propensity & Tree Based Models
Utilising Analytics to target Marketing Effort
APDS propensity models helped increase card spend by 15% in 2 months for a large Middle East issuer, see how we could help your organisation
Propensity Modelling 101, click to download
Bureau Scoring / Analytics
Does your bureau / CRA utilise proprietary scores from a global provider of scores?
​
Are the provided scores that are 'black box' in nature?
​
Do your customers need to understand the components / characteristics that constitute the bureau score for their Model Risk Compliance?
​
Do you want to be able to better support your clients when using your scores, through better model understanding?
​
APDS can build any CRA a transparent score with each characteristic available for explanation
Modelling with No Data
The Product Journey from No Data to Advanced Decisioning
Generic Models to Empirical Scoring: -
A generic or expert model (crafted by the our consultant's experience) is a tool that can be used when an organisation is new to a particular market and does not have relevant data to produce a tailored model for the product launch. The general development phase are
-
Information Gathering
-
Development of Characteristic Menu
-
Application to Market
-
Credit Policy Rules Developed
-
Volume based cut-offs
-
Implementation and data collect
-
Model Enhancement
After 6 months a monitoring pack is introduced, after 12 months the model may statistically fine-tuned (volume dependent).
Expert or Generic Model Build Approach
Regulatory Models
Compliance through Analytics
Over recent years models that predict a bank's capital and provisioning requirements have become increasingly important to both financial services providers as national and international regulators have become ever more vigilant to systemic risks within financial markets.
​
The main areas of focus have been capital (starting in the 2000s with the Basel Accords, accelerating after the Global Financial Crisis of 2008/9 and the introduction of global accounting standards in the 2010s.The key components of the regulations have been PD, EAD, LGD and Stress Testing models (ensuring the components can be adjusted for a range of downturn scenarios).
APDS can help your organisation build or enhance your regulatory models (quickly and efficiently), covering PD, EAD, LGD, macroeconomic models for stress testing and scenario planning & staging analytics
Monitoring, Validation & Model Risk
Model Monitoring - Confirming that your models are still working
Model Monitoring helps to determine when the operational or regulatory models deteriorate to a point that they cannot be used for the reason they were originally deployed. Monitoring also helps to determine how model performance dynamically changes of time.
Where models start to underperform the monitoring pack looks to determine the reasons why.
Key metrics
-
DiscoveryGini Co-Efficient / KS Test
-
Characteristic Misalignment
-
Population Stability at overall and characteristic levels
-
Rank Ordering
-
Override Reason / Policy Rules Review
-
TTD Tracking
​
Independent Model Validation
Independent Models Validation using a proven & repeatable framework
-
Discovery
-
Gap Identification
-
Data Audit - Confirmation of data being fit for purpose
-
Quantitative Validation - Replication of Results (Gini, KS, Predicted vs Actual, PD Grade Ratings etc.), Model Selection amongst Alternatives
-
Qualitative Validation - Confirming Modelling Assumptions, Model Selection, Post Model Adjustments
-
Report & Recommend
​
Model Risk
The Framework aims to offer continuous stakeholder engagement, transparent knowledge and no validation surprises​
Model Risk
As the use of models across the bank becomes more widespread covering almost every possible decision regulators have become increasingly about about the impact of using the models incorrectly or whether there are error within the modelling process or lifecycle.
​
Some regulators have introduced regulations to limit Model Risk, APDS can help banks comply or prepare to comply
Open Banking Analytics
Open Banking data is the next revolution to hit retail banking, allowing (with customer consent) participating banks to see a customer's entire banking relationships across multiple providers. This will enable forward thinking innovative organisations to steal a march on their competitors through the design of processes and procedure that helps to identify product need, assess the risk prior to application and provide tailored solutions to future customer needs.
​
What is needed is an comprehensive model infrastructure that takes the data, engineers features and builds the models that enable innovative bank decisioning.
​
APDS has a background in both risk and marketing models, therefore we feel ideally placed to help you exploit the opportunities that will arise.
Innovation - Alternative Data Scoring for Financial Inclusion & Economic Growth
70% of consumers in South East Asia are either unbanked or underbanked, by using alternative datasources to assess credit a Bank or MFI could steal a march on it's competitors
​
Whereas in Sub-Saharan Africa 51% do not have access to traditional banking services, however the introduction of Mobile Money has lift Financial Inclusion
Combining credit industry expertise with innovative datasources to drive credit business and economic growth and empowerment
Cambodia, Vietnam & Indonesia
Scoring for Financial Inclusion
Generic or Expert Trad Data Score
AlternativeData
Score
Model Use, Monitoring and RefinementÂ
towards Empirical Models and expansion to other South East Asia Countries
Combining Numerous Datasources to tackle
financial inclusion in Emerging Markets
We are working in South East Asia
(Vietnam, Cambodia (pilot implementation running with 6 lenders), Philippines and Indonesia (mobile scoring implemented) etc.) to drive economic growth and lift financial inclusion, the platform could be extended to African markets as well.
In many developing or emerging markets there are large segments of the population that have limited access to financial services, limiting opportunity and constraints potential economic growth. The limited access also makes it difficult for bank to assess credit worthiness and therefore cannot lend to potential.
​
There may be limited access to the banking but the consumers do generate data through the use of technology, which can be harnessed to assess credit applications.
​
APDS and it's partners are innovating to combine 30+ years of credit scoring experience with the newer datasets to offer banks a way to assess credit.
​
Training
Model Development Training
Train your new analysts in the the ways of Credit Risk Modelling
​
Build your own innovative models (powered by Paragon's Modeller software), with class leading strategies with detailed hands-on training by APDS
​
(Tailored Training with a 12 month mentor programme for all attendees is available for your Organisation)
​
See next page for our Training and Mentoring Programme
​
​
Edinburgh Credit Conference 2023 - Identifying Model Risks during Model Development
Meet the Team
Matthew Freeman has thirty plus years retail financial services industry experience, predominantly within the analytical risk management sphere, developing all manner of risk and marketing models, covering diverse geographies ranging from the United Kingdom and Western Europe, throughout the Middle East and India, China, Asia (North and South East) , Australasia (Australia and New Zealand)) and within emerging markets (such as Mongolia and Cambodia).
​
From a regulatory perspective Matthew has led teams (and built) all components of the AIRB and IFRS9 arenas in Europe, the Middle East and Asia (South East Asia, China and South Korea).
​
From an innovation perspective Matthew is leading the way in the use of alternative data for credit scoring / financial inclusion.
​
​