Abstract: In the option pricing literature, closed-form pricing formulas offer many advantages, but very few solutions are available. Among models that can incorporate the critically important stylized fact of stochastic volatility, the only known reliable solution for European options is the square root model in Heston (1993). Heston and Nandi (2000) offer a discrete-time alternative, but this is a GARCH-type model which does not feature stochastic volatility. We propose a new closed-form discrete-time option pricing model with stochastic volatility. The model is straightforward to implement. We estimate it using (jointly) a long historical time series of index returns and large option panels with various moneyness and maturities. The model vastly outperforms the existing discrete-time Heston-Nandi benchmark and slightly improves on the continuous-time benchmark. The model-implied pricing kernel and risk premiums are very plausible. The newly proposed pricing formula can be used to implement various extensions of the model.
Discussant: Bjorn Eraker, University of Wisconsin-Madison
Abstract: This paper examines the information content of news media for the cross-section of expected equity option returns. Applying various machine learning methods, we derive text-based signals from news articles on publicly traded companies that strongly forecast their delta-hedged option returns. The option return predictability is robust to variations in methodology and remains significant after controlling for existing predictors. We propose a text-based method to understand the underlying sources of our textual predictors. We find that the predictive power of the textual predictors stems from a composite effect, with future implied volatility changes being the most decisive, alongside significant contributions of various other option return determinants. Our study highlights the importance of analyzing text data using machine learning approaches to forecast option returns.
Discussant: Christopher Jones, University of Southern California
Abstract: On-Chain options refer to option contracts, that are traded directly on a Decentralized Exchange on the Ethereum blockchain. We explain the functioning of this new market form, so-called automated market making for options trading, and report a novel set of stylized facts. We provide a comprehensive analysis of On-Chain options and compare their attributes to their Off-Chain counterparts on centralized exchanges. We identify an On-Chain risk premium stemming from the positive disparity in implied volatility between On-Chain and Off-Chain options, attributing it to factors like the complex On-Chain fee structure, trading volume, and net demand pressure.
Discussant: Hui Chen, Massachusetts Institute of Technology
Dong Lou, London School of Economics and Political Science
Ke Tang, Tsinghua University
Abstract: We propose a novel measure, dubbed “relative basis,” to better capture the commodity convenience yield. Our measure is the difference between the traditional near-term basis and a similarly defined distant basis. This simple differencing purges out persistent commodity characteristics in traditional basis, such as storage and financing costs. Relative basis is closely tied to changes in physical inventories and dominates traditional basis in forecasting commodity futures returns. In contrast, relative basis does not forecast the returns of financial futures, which are not subject to inventory constraints. Our results provide new insights into the well-known relation between basis and expected futures returns.
Discussant: William Diamond, University of Pennsylvania
