Yaxiong Zeng

Northwestern University, Industrial Engineering & Management Sciences, Graduate Student

Followers

Following

Public Views

An enthusiast in machine learning applications to finance.
Phone: 8478686689
Address: 225 N Columbus Dr Unit 5608, Chicago, IL 60601

less

Guillermo Santamaria

Mazdak Nik Bakht

Concordia University (Canada)

Lucas Lai

George Brown College

Kiho Jeong

Kyungpook National University

Syracuse University

University of Madras chennai

Alan Liew

Interests

Uploads

Papers by Yaxiong Zeng

Online Adaptive Machine Learning Based Algorithm for Implied Volatility Surface Modeling

In this work, we design a machine learning based method, online adaptive primal support vector re... more In this work, we design a machine learning based method, online adaptive primal support vector regression (SVR), to model the implied volatility surface (IVS). The algorithm proposed is the first derivation and implementation of an online primal kernel SVR. It features enhancements that allow efficient online adaptive learning by embedding the idea of local fitness and budget maintenance to dynamically update support vectors upon pattern drifts. For algorithm acceleration, we implement its most computationally intensive parts in a Field Programmable Gate Arrays hardware, where a 132x speedup over CPU is achieved during online prediction. Using intraday tick data from the E-mini S&P 500 options market, we show that the Gaussian kernel outperforms the linear kernel in regulating the size of support vectors, and that our empirical IVS algorithm beats two competing online methods with regards to model complexity and regression errors (the mean absolute percentage error of our algorithm ...

Download

Machine Learning in Option Markets

Portfolio optimization for American options

Journal of Computational Finance, 2018

American options allow early exercise, which yields an additional challenge – besides the weights... more American options allow early exercise, which yields an additional challenge – besides the weights of each option – when optimizing a portfolio of American options. In this work, we construct strategies for an American option portfolio by exercising options at optimal timings with optimal weights determined concurrently. To model such portfolios, a reinforcement learning (Q-learning) algorithm is proposed, combining an iterative progressive hedging method and a quadratic approximation to Q-values by regression. By means of Monte Carlo simulation and empirical experiments, using data from the SPY options market, we evaluate the quality of our algorithms and examine their performance under various investment assumptions, such as different portfolio settings and distributions of the underlying asset returns. With discretized timings, our strategies work better in a relatively long time horizon and when the portfolio is hedged using the underlying instrument. Due to the highly leveraged and risky nature of our strategies, overly risk-averse investors are proved unsuitable for such investment opportunities.

Distributed solar renewable generation: Option contracts with renewable energy credit uncertainty

Energy Economics, 2015

Solar energy is rapidly emerging thanks to the decreasing installation cost of solar panels and t... more Solar energy is rapidly emerging thanks to the decreasing installation cost of solar panels and the renewable portfolio standard imposed by state governments, which gave birth to the Renewable Energy Credit (REC) and the Alternative Compliance Payment (ACP). To make profits from the REC market in addition to reduced energy costs, more and more home and business owners choose to install solar panels. Recently, third-party financing has become a common practice in solar panel investments. We discuss optimal timing for the host to potentially buy back the solar panels after being installed for a period of time and how to incorporate the optimal timing into a power purchase agreement between the host and the third-party developer. Because the REC price is a major source of uncertainty and also due to the ACP capping the REC price, we first propose a REC price forecasting model that specifically considers the ACP values. Then by a modified real option structure, we model the buyback contract as a real option and solve it with an approximate dynamic program based Monte Carlo simulation method. We find that as the ACP value increases, the value of the buyback option also increases under optimal timing. The method used does not only apply to solar projects but also to other distributed renewable projects that are third-party financed, such as wind generations.

Online adaptive machine learning based algorithm for implied volatility surface modeling

Knowledge-Based Systems, 2019

In this work, we design a machine learning based methodonline adaptive primal support vector regr... more In this work, we design a machine learning based methodonline adaptive primal support vector regression (SVR)to model the implied volatility surface (IVS). The algorithm proposed is the first derivation and implementation of an online primal kernel SVR. It features enhancements that allow efficient online adaptive learning by embedding the idea of local fitness and budget maintenance to dynamically update support vectors upon pattern drifts. For algorithm acceleration, we implement its most computationally intensive parts in a Field Programmable Gate Arrays hardware, where a 132x speedup over CPU is achieved during online prediction. Using intraday tick data from the E-mini S&P 500 options market, we show that the Gaussian kernel outperforms the linear kernel in regulating the size of support vectors, and that our empirical IVS algorithm beats two competing online methods with regards to model complexity and regression errors (the mean absolute percentage error of our algorithm is up to 13%). Best results are obtained at the center of the IVS grid due to its larger number of adjacent support vectors than the edges of the grid. Sensitivity analysis is also presented to demonstrate how hyper parameters affect the error rates and model complexity.

Download

Online Adaptive Machine Learning Based Algorithm for Implied Volatility Surface Modeling

Knowledge-Based Systems, 2019

In this work, we design a machine learning based method – online adaptive primal support vector r... more In this work, we design a machine learning based method – online adaptive primal support vector regression (SVR) – to model the implied volatility surface (IVS). The algorithm proposed is the first derivation and implementation of an online primal kernel SVR. It features enhancements that allow efficient online adaptive learning by embedding the idea of local fitness and budget maintenance to dynamically update support vectors upon pattern drifts. For algorithm acceleration, we implement its most computationally intensive parts in a Field Programmable Gate Arrays hardware, where a 132x speedup over CPU is achieved during online prediction. Using intraday tick data from the E-mini S&P 500 options market, we show that the Gaussian kernel outperforms the linear kernel in regulating the size of support vectors, and that our empirical IVS algorithm beats two competing online methods with regards to model complexity and regression errors (the mean absolute percentage error of our algorithm is up to 13%). Best results are obtained at the center of the IVS grid due to its larger number of adjacent support vectors than the edges of the grid. Sensitivity analysis is also presented to demonstrate how hyper parameters affect the error rates and model complexity.

Download

Portfolio Optimization for American Options

Journal of Computational Finance, 2018

American options allow early exercise, which yields an additional challenge when optimizing a por... more American options allow early exercise, which yields an additional challenge when optimizing a portfolio of American options, besides the weights of each option. In this work, we construct strategies for an American option portfolio by exercising options at optimal timings with optimal weights determined concurrently. To model such portfolios, a reinforcement learning (Q-learning) algorithm is proposed, combining an iterative progressive hedging method and a quadratic approximation to Q-values by regression. By means of Monte Carlo simulation and empirical experiments using data from the SPY options market we evaluate the quality of our algorithms and examine their performance under various investment assumptions, such as different portfolio settings and distributions of the underlying asset returns. With discretized timings, our strategies work better in a relatively long time horizon and when the portfolio is hedged using the underlying instrument. Due to the highly-leveraged and risky nature of our strategies, overly risk-averse investors are proved not suitable for such investment opportunities.

Download