Response Keynotes

• Monte Carlo methods need massive computational power and resources (high computational resources i.e. hardware risks remaining unused outside of active trading hours)
• Monte Carlo methods have slow to convergence rate.
• Monte Carlo methods are also unable to re-use previous computations: even small perturbations in one input requires starting from scratch.
• Monte Carlo methods unable to generate real-time analytics about asset dynamics and pricing behavior.

Response Keynotes

• ML algorithms are well-capable of transfer learning, wherein a model (or part of it) is re-trained for an application similar, but not identical to, its original purpose. A ML model can potentially be trained to approximate valuation for likely input parameter ranges on a single date — and recalibrated overnight.
• ML algorithms can be designed to dynamically auto-parameterize and auto-calibrate with the change in market data.
• ML brings in highly-potent computational functionalities like automatic differentiation (AD) found in most ML programming libraries. AD can deliver accurate model partial derivatives at little to no extra computational cost. Not only can a well-trained ML model then potentially value derivatives fast and accurately, it can also bypass FDM in evaluating the model input sensitivities for approximating true model sensitivities.

Response Keynotes

• Support Vector Machine (SVM) used to predict the outcome of exotic options. Like the case of double no-touch is a binary option with a constant payout earned if and only if the underlying asset price (mostly Forex) remains between a predefined lower and upper bound until expiration.
• Given the not continuous future payoff specified as “all-or-nothing” the SVM can be used to separate the two classes corresponding to the binary option outcome. As these classes are not linearly separable, one needs a kernel to predict the outcome. The features selected to train the model could be the average directional index and the ratio between realized volatility over implied volatility

Response Keynotes

• Structured products like Phoenix Autocalls which are traded over-the-counter (OTC) with institutional and private investors as part of larger structured products can be priced using Deep Learning algorithms like Neural Networks.
• Phoenix Autocalls depend on the price movements of multiple other financial assets in complex ways hence requires huge number of features and parameters to develop a pricing model. Further, being OTC, the payoff can be specialized in various different ways, The first-order sensitivities of these derivatives are evaluated for risk- management purposes, with sensitivities to the spot price (current price) and implied volatility of each underlying instruments particularly important; as well as sensitivity to time, measured by the interday (overnight) change in contract value. Present (theoretical) value and sensitivities to underlying asset price & volatility are continuously updated during market hours, requiring both fast and accurate valuation at a level not possible using Monte Carlo, but theoretically possible by deep learning.

Response Keynotes

• Due to the variety of hedge fund—and, therefore, investing strategies—it can be hard for investors to classify such investment vehicles. Moreover, hedge funds tend to reveal less information than other type of funds as they do not fall under the same disclosure requirements.
• To classify hedge funds, predefined classes would not be able to manage correctly future type of hedge funds. Hence, ML clustering methods, such as k-means, have been used to overcome cluster hedge funds by strategies like Event- Drive, Macro, Global-Arbitrage etc. The features considered are based on available characteristics of hedge funds, such as asset classes, size, fees, leverage and liquidity.