Why your credit models should no longer be built from scratch

Why your credit models should no longer be built from scratch

As we have outlined before in this blog, technology is already revolutionizing how investment, trading, research and risk management professionals across credit markets are accessing and scaling data for their end-to-end credit analyses.  

As data delivery and model building becomes fully automated, the “traditional" manual credit research process is becoming slow and outdated. As an ex-credit analyst and portfolio manager, I acknowledge that the pace and scale of technological evolution in our market can be unsettling. Technology is not going to replace human analysts - critical analysis of a credit is still where the real value lies - but technology can help that critical analysis shine by streamlining and expediting many of the time-consuming steps.

It was this idea that drove Cognitive Credit to build its own machine-learning technology specifically to translate those efficiencies to specialized data and analytics in credit markets. Today, it is clear that those already leveraging the benefits of data technology and automation are gaining a competitive advantage over their peers. So, what does the “traditional” credit analysis process now look like for them?

Based on my conversations with our clients, I outline the essential data-driven steps to a comprehensive credit analysis process that can leverage the benefits of technology and automation. In each step, I have used Cognitive Credit’s capabilities to illustrate how each of these steps is reduced from hours of manual work to just a few clicks.

Step 0: Screening for New Investment Opportunities

Screening for new investment or trading opportunities is an essential workflow in any credit or sector analysis process. It’s the starting point for any analyst, but often hard to do because data is not centralized, standardized or readily available.

Through our web application, our users save valuable time during this exercise via Comparables views, offering individual screens for Company, Sector and Markets. These screens give a view of our entire coverage universe, allowing users to easily compare issuers and sectors across a comprehensive list of fundamental and market data points to quickly hone in on mispriced securities.

 

 

Step 1: The First Cut of the Financials

The ‘First Cut’ of an issuer’s financials is a tricky process. Traditionally, there has been no good way to look at summary financials and trends across multiple periods without knowing if there is a tangible investment or trading opportunity at the end of it. Often, it’s a lot of frontloaded effort for no return. Hence the colloquial reference to a ‘quick and dirty look’.

With Cognitive Credit, the First Cut is a ‘quick and clean look’ facilitated by our auto-generated Memo feature. It’s literally a financial 1-pager for any issuer, summarizing all the essential fundamental data into a PDF in a single-click. It’s an enormous time-saver, and allows you to conduct ‘First Cuts’ in multiple issuers without much effort.

 

 

Step 2: Access & Documentation

Getting all the earnings reports, presentations, prospectuses, and sustainability reports together in one folder is a necessary but time consuming administrative process. For the rapidly increasing universe of private issuers, this process goes from time consuming to painstaking with the added steps of getting access to issuer portals or third-party data sites. And not to mention logging in again every time a new document is released or updated.

We remove all this hassle by hosting all the relevant documents used during our build process in a centralized document library; and because it has been fully processed for a model build, every document is fully auditable - allowing you to instantly search for any words or references across the entire collection of documents.

 

 

Step 3: Detailed Financial Model (via App & Excel Add-in)

This is the meat and potatoes of all credit analysis. Model building can take a long time to perfect as analysts manually enter or copy numbers from PDFs into their financial model template. Even the most minor typo can set you back significantly as you search through documents for the source of the error or set up balancing checks in the model - let alone major historical restatements!

Our technology reduces this days-long process into mere minutes - less than 15 minutes for Cognitive Credit users. In just a few clicks, our clients have access to the full historical financial model as reported by the company with 99.99% accuracy. Furthermore, these models also automatically update and flag all historical restatements or material changes to reporting, whilst providing full auditability between the original and restated number - automating away one of the most laborious tasks for credit analysts.

 

 

Step 4: Sensitivities & Forecasting

Building a forecast model for an issuer is also necessary for comprehensive credit analysis. As well as a current view, analysts need to have a forward looking view on every issuer under their coverage, and the ability to run sensitivities when there is a relevant news event or profit warning.

As we already have a full model with structured and tagged datapoints, our technology automates away the time spent building assumption sets with auto-generated integrated forecast models. With a Quarterly or Annual template available, each includes pre-populate model drivers for your analysis that can be easily customized to suit your needs. Once you have made your changes, you can save this Forecast model in-app, and revisit at a later date to instantly update with newly reported numbers as well as compare your forecast vs. actuals in each period.

 

 

Step 5: One Pager for Presentations

When it comes to discussing an investment or trading idea internally, every analyst has their 1-pager summary format built into their individual models for easy copy-paste into a word or pdf document. During a time crunch, preparing a 1 pager for one or multiple issuers can be stressful - and from the Cognitive Credit perspective - completely avoidable.

In addition to the Memo feature available in our web-application, our Excel Add-in is a powerful engine for your already built Excel models. With the Excel Add-in, users can recreate their chosen template format and simply change the ticker to switch between issuers to translate the speed and accuracy of the financial model into your presentation.

If your presentation format includes visual representation of underlying fundamental trends, our application makes it possible to simply select and drag rows, columns or areas to create instant and presentation ready charts.

 

 

Step 6: Earnings Reviews & New Issue Analysis

Earnings Season and New Issues can make certain weeks of the year incredibly heavy for analysts, PMs and Traders. The time required to manage risks or opportunities related to earnings headlines and evaluate new investment opportunities in the primary market leaves a very short working window. As a result, participants across the credit markets are forced to prioritize select names or drive decisions on incomplete information during these periods - often missing out on potential opportunities or sleepwalking into risk.

Introducing technology here eliminates the need for prioritization. For example, Cognitive Credit releases full model updates along with real-time earnings summaries delivered via email within 15 minutes after an issuer reports. For New Issues, the turnaround time is within 4 hours of receiving the OM and related documents. The result is that you get more time for critical analyses as well as greater accuracy and reach within your immediate coverage universe and the larger market during these busy periods. 

 

 

Cognitive Credit: Your own credit view, faster.

At Cognitive Credit, we are at the forefront of the technological revolution in credit markets. Our machine reading and extraction technology delivers extremely fast, accurate, and fully auditable data output, meaning no more offshore teams that come with human delays and errors.

Our technology also gives you operating leverage, allowing you to scale your issuer coverage in multiples. All our data is structured automatically, allowing for tremendous flexibility on how you use and manipulate the data without disrupting continuity.

If you’d like to see how our technology can supercharge your credit analysis process, request a demo today.