Factor Risk Model Correction for Missing Portfolio Risk Factors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current factor risk models are inadequate in accurately estimating portfolio risk due to inherent modeling errors and inability to capture all variance/covariance matrices, leading to underestimated risk predictions and suboptimal portfolio construction.
Innovation Solution
The solution involves modifying factor risk models by identifying and incorporating 'missing' or unspecified factors that account for modeling errors, using mathematical properties of the exposure matrix to adjust risk estimates based on portfolio holdings and historical performance, and implementing a computer-based method to calculate adjusted risk estimates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If factor risk models are used to estimate portfolio risk, then risk estimation can be performed efficiently, but the models inherently produce modeling error and underestimate risk
Solution Approach 1:
The patent introduces an intermediary correction term that mediates between the factor risk model estimate and the true portfolio risk. This correction term, derived from the exposure matrix properties, acts as a bridge that compensates for the modeling error without requiring complete reformulation of the risk model, thus maintaining efficiency while improving accuracy.
Solution Approach 2:
The patent modifies the risk estimation by changing the parameters used in the factor risk model. Specifically, it adjusts the risk estimate by incorporating the correction term that depends on the portfolio weights and the exposure matrix, thereby transforming the inadequate model output into a more accurate risk measurement.
2Device complexity
If traditional factor risk models are used, then computational complexity is reduced, but modeling errors lead to suboptimal portfolio construction
Solution Approach 1:
The patent segments the risk estimation process into two parts: the original factor risk model component and the correction term component. This segmentation allows the system to maintain the simplicity of the factor model while adding a separate, targeted correction that improves reliability without increasing overall model complexity significantly.
Solution Approach 2:
Rather than completely reformulating the risk model to account for all sources of error, the patent applies a partial correction that targets the specific modeling error arising from the exposure matrix. This partial action is sufficient to improve portfolio construction reliability without the excessive complexity of a complete model redesign.
3Device complexity
If factor risk models with fixed factors are used, then the model structure remains simple, but they cannot capture all variance and covariance information
Solution Approach 1:
The patent addresses the information loss by adding another dimension to the risk estimation. Instead of expanding the factor space horizontally (adding more factors), it adds a vertical dimension through the correction term that captures the missing variance-covariance information orthogonal to the existing factor space.
Data Source
AI summary
Techniques for using factor risk models to more accurately estimate the risk or active risk of an investment portfolio are disclosed. Inherent “modeling error” in factor risk models is identified and compensated for. One or more factors are added to compensate for factors that are unspecified or unattributed in the original factor risk model and which lead to modeling error. The approach can be used with a variety of different factor risk models, and for a variety of securities. Knowledge of the risk associated with modeling error can be utilized when estimating risk or active risk using factor risk models or when constructing optimal portfolios by mean-variance optimization or other portfolio construction strategies using factor risk models.