In practical applications, and especially in high dimensional models, it is common that not all covariances are known due to, for example, data limitations. In such situations, the covariance matrix must be completed in order to use the statisti-cal or stochastic model at hand. The already specified covariances generically imply a dependence between variables between which the covariance is unspecified. A criterion that can then be used to complete the covariance matrix is to introduce as little extra dependence between variables as possible which is equivalent with entropy maximization. Such completion problems can be solved with global optimization al-gorithms which unfortunately are slow for large matrices. And thus, this makes global optimization less suitable for applications in which the computation time is important. Another difficulty is that the initially specified covariances can be inconsistent in the sense that no valid completion exists. In this paper, we present a heuristic for complet-ing partially specified covariance (and correlation) matrices that is: fast (also for large matrices), tries to introduce as little extra dependence between variables as possible and also corrects for inconsistencies in the initially specified covariances. Finally, we present error measures for the performance of the heuristic.
Related Insights
-
18 November 2024Ortec Finance appoints Koen de Reus to Head of Client Servicing, Climate & ESG SolutionsKoen de Reus will lead a team of climate risk professionals to help financial institutions measure, manage, and monitor their portfolio’s climate risk exposure.
-
07 November 2024Insights from the FCA’s “Dear CEO” letter of October 7, 2024: What advisers need to knowThe Financial Conduct Authority (FCA) continues to stress the importance of firms prioritizing consumer needs and delivering good outcomes.
-
07 November 2024Exciting updates for our international Client Conference formatAnnouncement of some updates regarding the Ortec Finance annual International Client Conference.
-
30 October 2024How AI can help manage insurance portfoliosLearn how our Scenario-Based Machine Learning (SBML) approach enables insurers to optimize complex objectives by training AI on thousands of simulated scenarios, uncovering high-performing portfolios traditional methods might miss. Discover how AI-driven scenario analysis is transforming insurance portfolio management.
-
29 October 2024Quarterly Outlook: Q3 2024The outlook is based on the Ortec Finance Economic Scenario Generator (ESG) and offers our perspective on recent developments in the economy and capital markets.
-
23 October 2024Quarterly Pensions Investments ReviewThe Quarterly Pensions Investments Review is a comparison in expected risk and investment return.
-
08 October 2024The Importance of Financial Planning: A Spotlight on World Financial Planning Day 2024As we celebrate World Financial Planning Day 2024, here is why financial planning is more crucial than ever.
-
01 October 2024Ortec Finance appoints Maurits van Joolingen to head Climate & ESG SolutionsMaurits van Joolingen will be responsible for Ortec Finance’s climate scenario analysis tool to support the risk management needs of financial institutions.
-
30 September 2024Retirement Planning Myths Debunked: What You Really Need to KnowNavigating retirement planning can feel overwhelming, especially with the myths and misconceptions that surround it.