Product Overview

Applications

TARA™’s aim is to optimize predictions of future collateral performance, by learning and tracking the factors that influence historical (and current) loan performance.

  • Unprecedented accuracy: TARA™ autonomously risk rates loans’ future performance with unparalleled accuracy and speed, without the distortion of bias.

  • Higher returns: TARA™ provides forward looking results, while accepting the users’ assumptions as well, to generate custom stresses, enabling users to buy higher yield bonds lower in the stack, with confidence.

  • AI powered: The continual iterative machine learning processallows TARA™ to produce clear, concise, and objective risk assessments effectively and effortlessly in real-time.

  • Maintain quality standards: During phases of staff turnover, TARA™ helps ensure credit work accuracy and high-quality standards.

  • Increase productivity: TARA™ enables teams to refocus and better prioritize by directing Human Resources to their highest and best use. Allowing teams to free up time for deeper credit analysis and or platform expansion translates into greater accuracy and alpha.

Machine Learning vs. Deep Learning

Basic Methodology

  • Machine learning is a branch of AI that gives computers the ability to learn and improve with experience, without explicitly being programmed, by using data and algorithms to imitate intelligent human behavior. Algorithms use historical and present data as inputsto predict new values.

    Deep Learning is a subset of machine learning that uses vastvolumes of data and complex algorithms to train a model, teaching computers to do what comes naturally to humans, to learn by example.

    The machine learning process looks for patterns in data to later make inferences based on the examples provided, allowing computers to learn autonomously without human intervention and adjusting actions accordingly.

Why AI?
The application of AI in investment analysis cannot be described as “top-down”, at least not in the same way as a man-made credit risk model, because ofthe sheer amount of granular information that it considers. Neither can it be described as “bottom-up” because, after all, it is built using data. It is in between the “bottom-up” and “top-down” approaches; it is “bottom-up” when set next to a man-made credit risk model, and “top-down” when set next to an investment analyst’s qualitative approach.

Given the effectiveness of TARA™ at leveraging available data to predict strength of underlying bond collateral, it should be expected that TARA™ will be able to build portfolios that outperform in the future, and that any deviance would be under pressure to be corrected (i.e., the bonds that TARA™ flags as containing higher risk collateral will underperform against bonds that TARA™ flags as lower collateral).
TARA™
The granular nature of AI will make it more dynamic and accurate relative to man-made credit risk models. Its accuracy will be driven by exponentially more data than a regular model, and as such has the capacity to speed up the analyst, focus analytical efforts, and fill in with greater reliability than a regular model. This allows the analyst to get on with more in-depth analysis with confidence.

In addition, AI has asuperior ability to draw inferences in data. It can mimic any model, decision tree,or mathematical functionknown to man, when or if it deems it necessary to do so. This ability gives it an inherent advantage over typical man-made models.

Use Cases

  • Portfolio Management / Trading

  • Credit Analysis

  • Risk and Surveillance

  • Marketing / Fundraising / Training

Commercial Real Estate(CRE/CMBS)


TARA™ is an AI built for the purpose of tracking tranche and collateral performance. Typical usage (by function) for teams using TARA™ can be broadly broken down as follows:     
Across all asset types, TARA ™ combines loan level attributes with Economic indicators.  Within each asset, there are specific nuances that are also considered and are described below.

Collateral fields

  • Investor Reporting Package history

  • Historical Loan Pricing

  • Reference Data

  • Both collateral and bond model

  • Index support files

Asset Class Overview - COVERAGE

Collateralized LoanObligations (CLOs)

Coverage             Debt                 Equity

USD                                  7,500                           2,000
EUR                                  1H 2025                    1H 2025
Private (estimate)              1,000                           250    
 

Commercial Real Estate(CRE/CMBS)

Coverage                                  Equity

Conduit                                                                  12,000
SASB                                                                       2,500
CRE CLO                                                                    500

Model Overview - METHODOLOGY

Objective


TARA™ is an AI built for the purpose of tracking tranche and collateral performance. Typical usage (by function) for teams using TARA™ can be broadly broken down as follows: