top of page

THE PATENTS

FinaTech Has Patented the Computational Technology Needed to Model Alternative Assets With Sufficient Accuracy to

Issue, Rate, and Trade Structured Securities

UNLOCK THE POWER OF PRECISION PREDICTIVE ANALYTICS IN YOUR NEXT STRUCTURED FUND

The Difference Between Structured PE Funds and Mortgage-Backed Securities

Modeling and predicting the performance of mortgage-backed securities (MBS) is very different than modeling and predicting the performance of structured Alternative Asset-Backed Securities (AABS). To issue and rate a mortgage-backed security, simple actuarial data can be applied to a pool of loans to see how well the pool will fare under a range of economic conditions. 

CRE.png

To accurately project or rate the performance of a portfolio of companies, however, one must assess how each company will perform under a wide range of inputs, which requires a much more sophisticated methodology.

Complex AABS Diagram.png

WHEN ARE YOU USING PATENTED TALP-DERIVED PREDICTIVE ANALYTICS?

When one develops a model to predict or manage the outcome of a portfolio of companies, one must derive the algorithms one is using from various inputs, which may include factors like how a particular marketing plan could improve a company's revenue, or how replacing a company's obsolete technology could cut expenses. When coded, these predictive analytics become if/then statements expressed through data transformation algorithms that add to the cyclomatic complexity of one's software.

 

FinaTech's patents apply as the cyclomatic complexity of software using derived predictive analytics and data transformation algorithms increases. One necessarily crosses certain thresholds of complexity when one models portfolios of equities with multiple investor classes, which is why structured equity funds need licensing while unstructured equity funds (in which all the investors share equally in the fund's return) and debt funds generally do not need licensing.

The bottom line is if one is originating, issuing, rating, investing in, trading, managing, or selling interests in a structured fund, trust, or security based on or collateralized by equity assets, and the cashflow is allocated to more than one class of investor, one is likely violating FinaTech's patents. The full specifics of FinaTech's patents are quite technical, of course, and extend well beyond the mere use of predictive analytics and data transformation algorithms. An overview of some of the more relevant technical details in FinaTech's patents is provided below.

DESCRIPTION OF FINATECH'S PATENTS 

The following patent descriptions detail how each of FinaTech's patents cover essential aspects of the precision predictive modeling needed to accurately issue, rate, trade and manage structured AABS. One can look up FinaTech's patents by using the United States Patent and Trade Offices' (USPTO) public search engine at https://ppubs.uspto.gov/pubwebapp/static/pages/ppubsbasic.html. After entering the patent number, the USPTO’s archived PDF for each patent is accessed by clicking search and scrolling. The first three patents listed below have been issued directly to FinaTech’s IP holding company. The next three patents are licensed exclusively by FinaTech for use in AABS.

 

 

1.   US Patent #11,687,328 - Method and System for Software Enhancement and Management

 

 

This patent demonstrates that software can be enhanced and/or managed with significant performance gains using the TALP-derived predictive analytics of that software and any interacting software system or inputs. The patent demonstrates that multi-variable attribute input/outputs can be directly converted into algorithms, making this an important patent in general in the computation and software space. The patent uses a more general form of TALP-derived predictive analytics than shown in the other patents that FinaTech licenses, which broadens the range of analytics covered.

 

Among other things, this patent shows:

 

  • How to use TALPs to derive predictive analytics directly from data transformation algorithms in general and specifically data transformation algorithms representing companies and assets.

  • How to use TALPs to perform complex multi-variable parallelization with pinpoint accuracy on data transformation algorithms representing companies and assets.

  • How to use TALPs to construct the time complexity of an algorithm for arbitrary data events.

  • How all input/output pairs such as assets, investment criteria, rating agency criteria, debt criteria, market conditions, changes in technology, economic conditions, and fund structure can be treated as non-software algorithms and deconstructed into TALPs through the use of inherent analytics for more accurate modeling, predictions, and asset management.

  • How TALPs of representative algorithms can be simulated and/or modeled using TALP-extracted analytics for selection, rating, merging, management, discretization, and optimized distribution.

 

This patent demonstrates how assets and their optimized output values can be distributed over multiple categories of partners, investors, lenders, and rating agencies. Whenever the projected returns from a complex model of assets, bonds, or prioritized securities accurately predict the actual returns from a fund, they are necessarily using the TALP-derived analytics of the vehicle. 

Patent 1.png

2. US Patent #11,861,336 - Software Systems and Methods for Multiple TALP Family Enhancement and Management

 

 

This patent shows:

 

  • How software, algorithms, and the relationships between input and output data pairs can be transformed into TALPs with extracted intrinsic predictive analytics to optimize the accuracy and predictability of a model.

  • How to use and derive intrinsic, multi-dimensional predictive analytics to transform input/output matrices and array pairs into TALPs (other patents focus on the transformation of scalar values).

  • How a simulation of the functional characteristics of an algorithm is equivalent to TALP-derived intrinsic analytics and can be used to group assets into families, as in portfolios or funds, and into cross-families, as in funds of funds.

  • How to use TALP-derived analytics to manage assets that have been grouped into families, as in portfolios or funds. and cross-families, as in funds of funds.

  • How to use TALP-derived intrinsic analytics to project distributions of cashflow and risk to multiple categories of creditors, LPs, and GPs, each with their own inherent acceptance and rating criteria.

  • How to use TALP-derived intrinsic analytics to group, analyze, and manage different types of input data, such as changing economic conditions, rating criteria, investor criteria, and management decisions for the aggregation and discretization of output data for grouped assets such as portfolios, funds, and funds of funds.

  • How to use interacting grouped TALP-derived intrinsic analytics to manage assets and debt in portfolios, funds, or funds of funds.

  • How to use TALP-derived intrinsic analytics with externally defined data such as economic conditions and rating criteria to optimize a fund’s structure in terms of risk and/or returns, and how to optimize risk and/or returns for a given structure.

 

As with FinaTech’s other patent IP, this patent shows that the more accurately a simulation or model mirrors reality, the closer those simulations and models reflect the TALP-derived intrinsic predictive analytics. It also shows that TALP-derived intrinsic analytics are agnostic with regard to the use of any standard hardware configuration, including stand-alone servers, centralized computing client servers, decentralized clouds, and decentralized ad hoc networks.

TALP Structures.png

3. US Patent #11,914,979 - Software Systems and Methods for Multiple TALP Family Enhancement and Management - Continuation

 

  • Continuation of US Patent Application #18/102,638

  • Issued Date: February 27, 2024

  • Link: Not yet issued by the USPTO

  • Issued to FinaTech’s IP holding company

 

This patent shows:

 

  • How to use TALP-derived intrinsic analytics to generate inherent algorithmic and source code outputs that optimize the performance of algorithms and source code.

  • How to use TALP-derived intrinsic analytics with input data to select assets, creditors, investors, and fund structure.

  • How to use TALP-derived intrinsic analytics to optimize grouped asset performance based on inputs such as economic conditions, rating criteria, investor criteria, market changes, management decisions, and fund structure.

  • How to use TALP-derived optimization for PE fund data, REIT data, risk and return data, capital call data, prioritized investment unit data, cash flow data, securities data, and interest data.

  • How to use TALP-derived intrinsic analytics to discretize and optimize output data from pooled assets for multiple investors with different goals.

  • How to use TALP-derived intrinsic analytics to model, simulate, and optimize asset data, cashflow data, economic data, market data, rating criteria data, management data, and fund structure data.

 

As with FinaTech’s other patent IP, this patent shows that the more accurately a simulation or model mirrors reality, the closer those simulations and models reflect TALP-derived intrinsic analytics. It also shows that TALP-derived intrinsic analytics are agnostic concerning the use of any standard hardware configuration, including stand-alone servers, centralized computing client servers, decentralized clouds, and decentralized ad hoc networks.

Patent Diagram 4.png

4. US Patent #10,496,514 - System and Method for Parallel Processing Prediction

 

 

Among its many disclosures, this patent shows how to:

 

  • Use curve fitting to generate polynomials that accurately predict reality.

  • Predict parallelized performance with mathematical precision.

  • Derive four key prediction polynomials from algorithms: speedup, time complexity, space complexity, and overhead complexity.

 

This patent discloses the true limits to parallel processing, using a mathematical derivation to:

  • Prove the scaling limits for parallel processing predicted by Amdahl’s Law are incorrect.

  • Prove the scaling limits for parallel processing as predicted by Gustafson’s Law, once proposed as a correction to Amdahl’s Law, only represents a subset of the true scaling potential for parallel processing.

  • Prove a new law for parallel processing, Howard’s Law, that accurately predicts the true scaling limits for any algorithm, showing how to achieve greater performance gains from parallel processing than ever before thought possible.

 

The patent also demonstrates:

 

  • How precision parallel processing necessarily relies on intrinsic properties derived from algorithms as prediction polynomials in order to achieve accuracy.

  • How to extract intrinsic analytics from input/output relationships, software code, and algorithms (without limitation to the use or means of extraction).

  • How to optimize non-parallelized software.

  • How to derive prediction polynomials for processing time, memory, and overhead.

  • How the determination of work (such as when projecting changes in asset value or cashflow), algorithmic timings (such as when projecting returns as a function of time), and algorithmic overhead (such as when determining an asset’s overhead) all necessarily rely on prediction polynomials derived from an algorithm’s intrinsic analytics.

 

This patent unlocks the means to more accurate curve fitting, showing that the ability to fit a curve formed by a set of input and associated output values is necessarily based on the intrinsic properties of algorithms in the environment being modeled. Therefore, the more accurate a model of a complex system is, the more it matches reality and the intrinsic analytics derived from input/output relationships. In other words, the more accurate a model, the closer it is to being indistinguishable from FinaTech’s patent IP.

5. US Patent #11,520,560 - Computer Processing and Outcome Prediction System and Method

 

 

This patent defines time-affecting linear pathways (“TALPs”) as a way to analyze and optimize software code and algorithms. TALPs play a key role in FinaTech’s patent portfolio. This patent expands on the concepts of US Patent #10,496,514, showing:

  • How to use input variable attribute values that define non-loop control conditions to decompose an algorithm or software code into TALPs.

  • How to eliminate temporal (workload) ambiguities in algorithms that execute on serial or parallel Turing machines through TALP decomposition and analysis.

  • How to use the input variable attribute values and their associated output values of TALPs to derive the intrinsic predictive analytics of an algorithm or software code.

  • How to use TALPs to ensure that derived intrinsic predictive analytics accurately reflect objective reality.

  • How TALP-derived intrinsic analytics can be applied to the outputs of a software code or algorithm.

  • How TALP-derived analytics can be applied to topological analysis to treat high dimensional algorithms as though they are linearly related.

  • How optimization methods, such as parallel processing (modeling multiple assets to project returns simultaneously for multiple investors), dataset curve fitting (determining when a fund will hit targets), and predicted data intercepts (determining the actions necessary to hit a target), can be derived from input/output relationships, software code, and algorithms using TALP decomposition and analysis.

 

 

6. US Patent #11,789,698 B2 - Computer Processing and Outcome Prediction System and Method

 

 

This patent shows:

 

  • How to derive TALP intrinsic analytics from a list of input/output data pairs using a self-learning search-and-compare curve-fitting model.

  • How to expand from a single input variable to multiple input variables used by a TALP to perform a curve fit, demonstrating a new advanced approach to non-linear multi-variable curve fitting that greatly reduces the computational overhead conventionally associated with partial differential equation curve-fitting solutions.

  • How to automatically select TALPs for analysis from various input/output value pairs, source code, or algorithms.

  • How to use TALPs to determine the original input values of an algorithm by observing the output values, demonstrating that accurate back tracing of values requires TALP-derived intrinsic analytics.

  • How to use TALPs to predict multi-algorithm race conditions like data collisions and other algorithmic interactions.

  • How to use TALPs to automate white-box and black-box testing of algorithms.

  • How to discretize an algorithm’s output values from TALP-derived intrinsic analytics.

  • How to use TALPs to convert standard algorithms, input/output dataset pairs, and software codes into quantum computable form.

 

TALP-derived intrinsic analytics are shown to be agnostic with regard to any standard hardware configuration, including stand-alone servers, centralized computing client-servers, decentralized clouds, and decentralized ad hoc networks.

bottom of page