Alex Khan, Clark Alexander Ph.D., and I completed our work on classical portfolio optimization. We can run classical jobs for a long time that optimize more than 32 assets at one time. We can run, brute force, 32 assets in a portfolio on an aged laptop, and can make progress towards completing 40 assets in one go. We stopped that run, which requires over 1 trillion portfolios analyzed, as our server was producing enough heat to keep our furnace redundant. We have developed a concept and associated workflow that allows us to analyze hundreds of assets and create superior portfolios.
We ran the Sharpe Ratio, and created our own 'proprietary' classical formulation which we call the Chicago Quantum Ratio. We achieve similar portfolio optimization results with cleaner math that should run better on a quantum computer.
Since April 10 we have been running different formulations of the problem (with 20 assets) as a QUBO on D-Wave Systems using solver DW_2000Q_5, which is a 2030 qubit system. We continue to run different formulations of the QUBO (or Binary Quadratic Model, BQM), to replicate the optimal portfolios as selected from using the Sharpe Ratio, and the Chicago Quantum Ratio.
- we started with one sample, random, portfolio and achieved acceptable results
- running a second sample, also random, that has significant negative correlations between stocks
- setting up a third random sample...
Our goal is to replicate or deliver equivalent results to our brute-force solvers, then...then increase the problem size that can be run on a quantum system.
As always, please reach out to us for more information, to work with us, or to join the team.
Clients are always welcome!
Thank you, Jeffrey Cohen, President US Advanced Computing Infrastructure, Inc., on behalf of our team.
Strategic IT Management Consultant with a strong interest in Quantum Computing. Consulting for 29 years and this looks as interesting as cloud computing was in 2010.