Risk Management with Quantum Computing. IonQ And GE Research Announce The Results Of Study

Risk Management With Quantum Computing. Ionq And Ge Research Announce The Results Of Study

After months of research, IonQ, and GE Research, a division of GE Electric, have announced the results of their experiment using quantum computers to model multivariable distributions in risk management. The work was done using IonQ Aria, IonQ’s 20AQ trapped ion quantum computer, and the results are remarkable. IonQ Aria is currently in private beta testing and will be available on Microsoft Azure later this year.

IonQ and GE Research trained quantum circuits with real-world data on historical indexes to forecast future performance. They used Copulas, a mathematical tool for simulating joint probability distributions.  Their quantum predictions occasionally outperformed classical copula modeling approaches, and they have the potential to identify higher probabilities of instability events occurring in many industries. The joint research work was accomplished using hybrid quantum computing, in which a classical computer handles some aspects of a problem while a quantum computer handles others. 

In a separate work that they conducted with a significant financial services company last year, they were able to model stock prices using IonQ Harmony, an 11-qubit quantum computer currently offered to customers on all three major cloud platforms. In that experiment, IonQ used two variables to model the likelihood of black swan events. They expanded that to three and four variables in various indexes in this new research with GE Research. Since hybrid quantum computing is dependent on optimization, IonQ developed an original type of optimization known as annealing training while working on the project. By creating more powerful quantum computers, they now have a set of techniques and tools that, if scaled up, will outperform traditional computing techniques. These experiment results prove that quantum copulas can outperform predictions made about sizable data sets using only classical computing equipment.

See More