Introducing Cognitive Credit AI for faster, smarter credit analysis | Read more
Book a demo

Markets

Accurate, comprehensive, on-demand credit market data

Machine-extracted, expertly verified fundamental data across US, European and emerging market bonds and loans.

Our coverage

Markets we cover

High Yield

1,600+ issuers

Structured fundamental data covering high yield bond issuers in US and European markets, as defined by the Markit iBoxx Liquid High Yield Indices.

Learn more

Investment Grade

1,100+ issuers

Structured fundamental data covering investment grade bond issuers in US and European markets, as defined by the Markit iBoxx Liquid Investment Grade Indices.

Learn more

Emerging Markets

100+ issuers

Structured fundamental data covering bond issuers in emerging markets, as defined by the J.P. Morgan CEMBI Broad Diversified Core Index.

Learn more

Leveraged Loans

1,100+ issuers

Structured fundamental data covering leveraged loan issuers in US and European markets as defined by the Morningstar Leveraged Loan Indices.

Learn more

New issues

Ramp up on new issuance faster than the competition.

Get a full document library and a ready-to-use credit model for inaugural issuers within six hours of deal announcement, and repeat issuers within minutes — only with Cognitive Credit.

INAUGURAL

<6hrs

Average model turnaround after inaugural new issue

REPEAT

<15mins

Average model turnaround for repeat issuers

How we deliver our data

1 of 5
Income statement illustration with fields highlighted
Cognitive Credit data extraction
Cognitive Credit data validation
Data displayed in Cognitive Credit web app with audit card
Data published to Cognitive Credit web app
Income statement illustration with fields highlighted

Step 1

Official company document is released

When new disclosure hits the market, our team sources the document directly from the issuer and submits it to our proprietary data extraction technology. This includes a mix of preliminary reports, earnings reports, presentations and prospectuses.
Cognitive Credit data extraction

Step 2

Raw data is automatically extracted

Using advanced character recognition and machine-learning technology, relevant data is identified and extracted from the document and classified into line items ahead of structuring. All data is extracted exactly as-reported. 

Cognitive Credit data validation

Step 3

Classified data is structured and validated

Our technology then algorithmically validates all extracted data points against known accounting principles and financial relationships within any historically reported data. This allows the data to be programmatically checked and enriched with any previously reported credit metrics. 

 

Data displayed in Cognitive Credit web app with audit card

Step 4

Our experts confirm model integrity

Once the data has been validated and structured into a model, our experts check all extracted data, derived values and associated formulae for any errors or inconsistencies. Restatements are also highlighted for full transparency. 

Data published to Cognitive Credit web app

Step 5

Model is published to Cognitive Credit

Once all checks are complete, the fully structured and ready-to-use model is published to our web app and made available for consumption via our API. On average, existing models are updated within 15 minutes of earnings release and new issues within six hours of deal announcement.

Comprehensive, accurate and fast

 

arrow-icon-svg

Sourced directly

 All data is sourced from official earnings reports, presentations and prospectuses

target-icon-svg

Unmatched accuracy

Data is identified, extracted and structured automatically by proprietary technology 

check-screen-icon-svg

Validated by technology

Every data point is algorithmically validated against accounting and financial relationships

magnifying-glass-icon-svg

Enriched automatically

Data gaps and accounting inconsistencies are resolved against the latest and historical reporting

check-user-icon-svg

Verified by experts

All outputs are verified by our team of analysts to ensure unparalleled data integrity

clock-icon-svg

Real-time updates

Existing models are refreshed within 15 mins (avg.) of earnings release. New issues added within six hours

Frequently asked questions

Where do you source your data from?

 All data on Cognitive Credit is sourced from official filings only. This includes a mix of Preliminaries, Prospectuses, Financial Reports and Presentations. Every document used to build our datasets is stored in our searchable document library.

How do you extract your data?

 We use a combination of advanced optical character recognition and machine-reading technology to automatically identify and extract data from official documents. Proprietary technology then classifies and structures this data into a dynamic credit model, with every datapoint linked to the exact page in a document from which it was extracted.

How do you identify restatements?

All restatements and inconsistencies in a dataset can be easily highlighted in-app for full transparency. Every restated or inconsistent value can also be audited to show previously reported figures and the calculations behind them.

Do you process restricted data?

 Yes - our technology can process restricted data. Using a simple document uploader, you can submit your restricted documents and our technology will automatically generate a dynamic credit model that reflects your documents and is only accessible by you.

How quickly do you turnaround models?

New coverage models are delivered within 24 hours of receiving documentation. Debut issuance models are delivered within 6 hours of announcement. New numbers on existing issuers are updated within 15 minutes (avg.) 

Market-leading fundamental data

Book a demo today

See how we can transform your research with a one-to-one demonstration tailored to your coverage requirements.