Portfolios up to 40% more efficient than the standard heuristic.

A provably optimal portfolio optimizer for swissQuant's institutional clients.

+40%

better risk–return than the standard heuristic

Proven

best answer, not an estimate

5 years

of real MSCI World data tested

Ready

for swissQuant's valuation pipeline

The problem

swissQuant's institutional clients need portfolios where no single asset adds too much risk. That rule makes the maths hard: standard solvers fail or time out, and heuristics return an answer with no guarantee of how good it is, which is hard to defend in front of a portfolio manager.

What we built

A custom optimizer that finds the provably best portfolio under that rule, plus two fast heuristics for when speed matters more than proof. Delivered as documented Python code, ready for swissQuant's valuation pipeline.

The result

On five years of real MSCI World data, it found portfolios up to 40% more risk–return efficient than the standard heuristic on 33-asset portfolios, each with a certificate that no better portfolio exists.

A problem like this?

Send it to us. Within days, you see a first version running on your own data.

Describe your problem

Next case~€1M: Truck loading that fills every truck, and never breaks a weight rule