Operational Research Assignment

Identify the Right Portfolio for Each Customer, While Addressing Compliance

Business Case Summary

The swissQuant Group develops and delivers intelligent technology products for Risk Modelling & Analytics, Trading & Risk Management and Hedging & Procurement.

In this business case; their primary objective is to implement a solution designed to alter the process of suggesting personalized investment portfolios for private banking customers.

This solution aims to integrate customer-specific objectives, risk profiles, and available assets while adhering to the stringent policies and regulations governing the banking industry.

Problem Description

Problem Description: Structuring a portfolio by formulating it as a mathematical optimization problem and optimizing it with respect to target function in order to maximize the portfolio’s rate of return subject to risk limits.

In this business case; solution provider Gurabi solves a mixed-integer quadratic problem (MIQP).

Portfolio optimization results in a MIQP as it is characterised by a quadratic objective function with 1,000 to 10,000 variables and almost as many additional linear conditions, and also some of the variables must only be represented as integer values.

Key Considerations to choose Gurabi’s Solutions: Better solver performance, higher speed, the ability to scale the project demands, stability and reliability in solving complex problems, better quality compared to its alternatives.

Solution Approach

Integration Process of Gurabi’s Solution:

  1. Creation of a customer profile

  2. Presenting opportunities and risks interactively in the profile, applying stress tests to understand bank risks

  3. Adding customer-specific restrictions and general bank rules and creating a diverse set of conditions

  4. Solving a mathematical optimization model, optimizing the portfolio to maximize return within risk limits.

  5. Presenting the optimized portfolio, featuring various investment strategies, risks, and benefits

Benefits

  1. Efficiency in Problem Solving: found solutions for 98% of feasible problems1 within 20 seconds, showcasing its efficiency in problem-solving.
  2. Better Performance compared to Competitor: The competitor could achieve solution times of less than 20 seconds for only 38% of the models tested.
  3. Optimal Solutions: found the optimum solution for 90% of the test problems, surpassing the competitor’s performance, which could only achieve this for 55% of the problems.
  4. Higher Returns: 1% higher returns compared to another commercial solver in portfolio optimization, In 7% of feasible problems

References

Business Case

Proven Techniques for Solving Financial Problems with Gurobi

Mixed Integer Programming Basics

    • A feasible problem is defined as a portfolio optimization with at least one solution that fulfills all restrictions.
    ↩︎