GARCH Models

Is the difference between deep hedging and delta hedging a statistical arbitrage? ArXiv ID: 2407.14736 “View on arXiv” Authors: Unknown Abstract The recent work of Horikawa and Nakagawa (2024) claims that under a complete market admitting statistical arbitrage, the difference between the hedging position provided by deep hedging and that of the replicating portfolio is a statistical arbitrage. This raises concerns as it entails that deep hedging can include a speculative component aimed simply at exploiting the structure of the risk measure guiding the hedging optimisation problem. We test whether such finding remains true in a GARCH-based market model, which is an illustrative case departing from complete market dynamics. We observe that the difference between deep hedging and delta hedging is a speculative overlay if the risk measure considered does not put sufficient relative weight on adverse outcomes. Nevertheless, a suitable choice of risk measure can prevent the deep hedging agent from engaging in speculation. ...

Volatility-based strategy on Chinese equity index ETF options ArXiv ID: 2403.00474 “View on arXiv” Authors: Unknown Abstract This study examines the performance of a volatility-based strategy using Chinese equity index ETF options. Initially successful, the strategy’s effectiveness waned post-2018. By integrating GARCH models for volatility forecasting, the strategy’s positions and exposures are dynamically adjusted. The results indicate that such an approach can enhance returns in volatile markets, suggesting potential for refined trading strategies in China’s evolving derivatives landscape. The research underscores the importance of adaptive strategies in capturing market opportunities amidst changing trading dynamics. ...

Forecasting Bitcoin Volatility: A Comparative Analysis of Volatility Approaches ArXiv ID: 2401.02049 “View on arXiv” Authors: Unknown Abstract This paper conducts an extensive analysis of Bitcoin return series, with a primary focus on three volatility metrics: historical volatility (calculated as the sample standard deviation), forecasted volatility (derived from GARCH-type models), and implied volatility (computed from the emerging Bitcoin options market). These measures of volatility serve as indicators of market expectations for conditional volatility and are compared to elucidate their differences and similarities. The central finding of this study underscores a notably high expected level of volatility, both on a daily and annual basis, across all the methodologies employed. However, it’s crucial to emphasize the potential challenges stemming from suboptimal liquidity in the Bitcoin options market. These liquidity constraints may lead to discrepancies in the computed values of implied volatility, particularly in scenarios involving extreme moneyness or maturity. This analysis provides valuable insights into Bitcoin’s volatility landscape, shedding light on the unique characteristics and dynamics of this cryptocurrency within the context of financial markets. ...

Comparing Deep Learning Models for the Task of Volatility Prediction Using Multivariate Data ArXiv ID: 2306.12446 “View on arXiv” Authors: Unknown Abstract This study aims to compare multiple deep learning-based forecasters for the task of predicting volatility using multivariate data. The paper evaluates a range of models, starting from simpler and shallower ones and progressing to deeper and more complex architectures. Additionally, the performance of these models is compared against naive predictions and variations of classical GARCH models. The prediction of volatility for five assets, namely S&P500, NASDAQ100, gold, silver, and oil, is specifically addressed using GARCH models, Multi-Layer Perceptrons, Recurrent Neural Networks, Temporal Convolutional Networks, and the Temporal Fusion Transformer. In the majority of cases, the Temporal Fusion Transformer, followed by variants of the Temporal Convolutional Network, outperformed classical approaches and shallow networks. These experiments were repeated, and the differences observed between the competing models were found to be statistically significant, thus providing strong encouragement for their practical application. ...