In the financial industry, common practice is to use something called hedge to reduce the risk of an existing position by taking an offsetting position in a different asset. For example, an investor may use a derivative such as a Put or Call to offset potential future losses by changes in the price of the underlying asset. While classic hedging algorithms may perform well in ideal markets, they may not perform well in real-world markets which can lead to account transaction fees, liquidity issues, trade restrictions and other issues.
In recent years, a framework called Deep Hedging has been developed to deal with this problem using an approach called reinforcement learning. This research was conducted by QC Equipment And JPMorgan Chase & Co. Global Center for Applied Technology Research explored using quantum deep learning algorithms to apply more efficient quantum reinforcement learning techniques for Deep Hedging. The potential benefit is being able to train a quantum neural network with fewer trainable parameters than the equivalent classical approach and increases the accuracy and trainability of models that will run on high-performance GPU hardware. The hardware platform used in this study is Quantinum H1-1 and H1-2 ion-trapped quantum processor.
March 30, 2023