Monte Carlo is synonymous with fast cars, fast money and – in financial circles – a painfully slow way to price derivatives.

Now banks are turning to machine learning to give their pricing models a turbo-boost.

The technique involves training so-called deep neural networks to approximate the results of Monte Carlo models without the need to run millions of simulations.

“A neural network approaches the price of your portfolio when you’re running a gigantic, complicated XVA

You cannot copy this content at this time. Please contact info@fx-markets.com find out more.

Source Link