DMC 2023 Programme
COMPLETE PROGRAMME | ||
ACADEMIC SCHEDULE | ||
DAY 1: PROGRAMME OVERVIEW, THURSDAY 7TH SEPTEMBER 2023 | ||
|
THURSDAY 08:45 to 09:00 | REGISTRATION & ARRIVAL COFFEE AND TEA | WF8 Lounge |
|
THURSDAY 09:00 to 09:30 | WELCOME AND INTRODUCTION: Bart Frijns, Open University | ROOM WF710 |
|
THURSDAY 09:30 to 10:00 | MORNING TEA BREAK | WF8 Lounge |
|
THURSDAY 10:00 to 12:00 | SESSION 1 | ROOM WF710 |
| CHAIRPERSON | Prasad Hegde, Auckland University of Technology | |
| PAPER | MEDIA EMOTION INTENSITY AND COMMODITY FUTURES PRICING | |
| Lina El-Jahel, University of Auckland | ||
| Yeguang Chi, University of Auckland. | ||
| Thanh Vu, University of Auckland | ||
| Abstract: | ||
| We investigate the role of media emotion in commodity futures pricing and propose a new factor, media emotion intensity, based on the proportion of emotional content relative to factual content. Our factor exhibits an annual premium of around 14% after we control for other commonly considered benchmark factors. The impact of media emotion is especially strong for commodities with low media coverage, high momentum, high basis-momentum, high hedging pressure, and backwardation. Media emotion intensity significantly predicts the cross-section of commodity futures return both at the portfolio level and the individual commodity level. Our simulated LASSO approach suggests that media emotion intensity is the most robust factor compared to other commonly considered benchmark factors. Furthermore, we investigate various risk channels that are potentially related to media emotion intensity and demonstrate that they cannot subsume the predictability of this media factor. | ||
| PRESENTER | Thanh Vu, The University of Auckland | |
| DISCUSSANT | Tingxi (Riven) Zhang, Curtin University | |
| PAPER | THE INFORMATION VALUE OF MEDIA COVERAGE OF MARKET VOLATILITY: A TEXTUAL ANALYSIS | |
| Ming-Hung Wu, Beijing Normal University | ||
| Wei-Che Tsai, National Sun Yat-sen University, and Risk and Insurance Research Center, National Chengchi University | ||
| Nai-Wen Cheng, The Taiwan Futures Exchange | ||
| Yi-Wei Chuang, National Kaohsiung University of Science and Technology | ||
| Abstract: | ||
| Our primary aim in this study is to measure media sentiment using textual analysis of news stories, blog posts, and discussion messages, before going on to explore the links with market sentiment based upon large-scale web data feeds and VIX futures returns. Our results reveal that whilst the sentiment index (calculated overnight) can indeed predict daily VIX futures returns, its predictive power is weakened by macroeconomic announcements. The sentiment effect is also found to be more pronounced on days with high numbers of postings, trading volume, volatility, and illiquidity. Following the strategies highlighted by the media sentiment index, our portfolio exhibits high performance, particularly when the analysis relates to news articles. These findings suggest that media sentiment contains the economic value in volatility trading. | ||
| PRESENTER | Yiwei Chuang, National Kaohsiung University of Science and Technology | |
| DISCUSSANT | Adrian Fernandez-Perez, Auckland University of Technology | |
| PAPER | SOCIAL MEDIA SENTIMENT, HERDING AND MARKET INFORMATIONAL EFFICIENCY | |
| Ni Yang, Auckland University of Technology | ||
| Adrian Fernandez-Perez, Auckland University of Technology | ||
| Ivan Indriawan, University of Adelaide | ||
| Abstract: | ||
| This study investigates the impact of social media sentiment on financial market informational efficiency. Specifically, we interpret the qualitative Twitter emotions into quantified social media sentiment and examine how Twitter Bullishness affects return autocorrelation and variance ratio in a high-frequency context. We find that a higher Twitter Bullishness increases the following day's intraday return autocorrelation and variance ratio, indicating that Twitter sentiment reduces the market informational efficiency. The relationship between Twitter Bullishness and market informational efficiency persists after controlling for lagged efficiency, contemporaneous returns, intraday realized volatility, trading volume, market depth and VIX. Our results withstand the choices of sentiment analysis approaches and intraday sentiment intervals. Furthermore, we assess the mechanism by which social media sentiment influences market quality. We find that the impact of Twitter Bullishness is due to herding behaviors among traders. A higher Twitter Bullishness is associated with a higher herding effect in the following day during trading hours, while herding does not cause subsequent relevant higher sentiment on social media contrariwise. This study shows that investors are not able to efficiently react to social media sentiment, which may exacerbate the effective dissemination of information, and worsen the informational efficiency. This creates higher levels of fractions and costs of trading at intraday level, decreasing the market information incorporation process. | ||
| PRESENTER | Ni Yang, Auckland University of Technology | |
| DISCUSSANT | Thanh Vu, The University of Auckland | |
| PAPER | NEWSWIRE TONE-OVERLAY COMMODITY PORTFOLIOS | |
| Adrian Fernandez-Perez, Auckland University of Technology | ||
| Ana-Maria Fuertes, University of London | ||
| Joëlle Miffre, Audencia Business School and Institute Louis Bachelier | ||
| Nan Zhao, Barclays Corporate and Investment Bank | ||
| Abstract: | ||
| We propose a method to overlay the tone of commodity-specific newswires upon the commodity characteristics traditionally used in long-short portfolio allocations. Implementing the tone-overlay strategy on 26 commodities generates substantial risk-adjusted profitability gains relative to the corresponding plain-vanilla traditional portfolios. Recession risk and limits-to-arbitrage risk emerge as key channels for the observed outperformance. The benefits of the tone-overlay tactical allocation are more pronounced when it focuses on very salient pessimistic or optimistic newswire tone in line with theories of limited investor attention. The tone-overlay approach is shown to be more effective than alternative approaches to embed newswire tone into traditional commodity allocations such as the equal-weight style-integration and double-sorting. | ||
| PRESENTER | Adrian Fernandez-Perez, Auckland University of Technology | |
| DISCUSSANT | Yiwei Chuang, National Kaohsiung University of Science and Technology | |
| THURSDAY 12:00 to 13:00 | LUNCH BREAK | WF8 Lounge |
| THURSDAY 13:00 to 15:00 | SESSION 2 | WF710 |
| CHAIRPERSON | Erik Schlogl, University of Technology Sydney | |
| PAPER | A VARIATIONAL FORMULATION OF EUROPEAN OPTION PRICES IN THE 1-HYPERGEOMETRIC STOCHASTIC VOLATILITY MODEL | |
| Wenjun Zhang, Auckland University of Technology | ||
| Abstract: | ||
| The paper proposes a variational analysis of the 1-Hypergeometric stochastic volatility model for pricing European options. The methodology involves the derivation of estimates of the weak solution in a weighted Sobolev space. The weight is closely related to the stochastic volatility dynamics of the model. The solution is further analysed using semigroup theory applied to the pricing operator. A full implementation of the model using the infinite element method is performed as well as a model calibration using a set of options to illustrate how the model works in practice. The analysis of the volatility distribution confirms the advantages of the model. | ||
| PRESENTER | Wenjun Zhang, Auckland University of Technology | |
| DISCUSSANT | Erik Schlogl, University of Technology Sydney | |
| PAPER | AN EMPIRICAL STUDY ON THE EARLY EXERCISE PREMIUM OF AMERICAN OPTIONS: EVIDENCE FROM OEX AND XEO OPTIONS | |
| Weihan Li, University of Otago | ||
| Jin E. Zhang, University of Otago | ||
| Xinfeng Ruan, Xi’an-Jiaotong Liverpool University | ||
| Pakorn Aschakulporn, University of Otago | ||
| Abstract: | ||
| Since the S&P 100 Index underlies both American (OEX) and European (XEO) options, the value of the early-exercise premium of American options can be directly observed. We find that the mid-quote of an XEO option can be higher than that of an otherwise identical OEX option, and liquidity can explain this overpricing phenomenon of European options. Our results show that illiquid options are significantly overpriced in the S&P 100 Index options market. This finding indicates that an illiquid option can be overvalued with a higher market offer price, which is the requirement of market makers for compensation to provide liquidity. | ||
| PRESENTER | Weihan Li, University of Otago | |
| DISCUSSANT | Guanglian Hu, University of Sydney | |
| PAPER | SOFR TERM STRUCTURE DYNAMICS DISCONTINUOUS SHORT RATES AND STOCHASTIC VOLATILITY FORWARD RATES | |
| Alan Brace, Financial Mathematics, Modelling and Analysis | ||
| Karol Gellert, University of Technology Sydney | ||
| Erik Schlogl, University of Technology Sydney, University of Cape Town, and University of Johannesburg | ||
| Abstract: | ||
| As more and more jurisdictions transition from LIBOR-type interest rate benchmarks to new risk-free rate (RFR) benchmarks based on overnight rates, such as SOFR in the US, it is important to adapt interest rate term structure models to reflect this. In particular, overnight rates are largely driven by monetary policy and thus display dynamics that are (at least to first order) piecewise constant between central bank rate decisions, while forward rates continue to evolve in a more diffusive fashion. We construct a tractable multifactor, stochastic volatility term structure model which incorporates these features. Calibrating to prices for options on SOFR futures, we achieve a good fit to the market across available maturities and strikes in a single, consistent model. The model also provides novel insights into SOFR term rate behaviour (and implied volatilities) within the SOFR term rate accrual periods, as well as into empirical mean reversion dynamics. | ||
| PRESENTER | Erik Schlogl, University of Technology Sydney | |
| DISCUSSANT | Weihan Li, University of Otago | |
| PAPER | VOLATILITY RISKS IMPLIED FROM SHORT-TERM VIX FUTURES | |
| Guanglian Hu, University of Sydney | ||
| Abstract: | ||
| We use a dynamic term structure model to extract latent volatility risk factors from short-term VIX futures. While the first factor, closely related to the level of volatility, does not contain predictive information about VIX futures returns, the second and third risk factors can significantly predict daily and weekly returns of VIX futures. The predictive power of the third volatility factor is particularly strong: It is robust to controlling for other known predictors, considering different VIX futures contracts and return calculation, and alternative methods for evaluating statistical significance. We find the third volatility factor captures both changes in risk and movements in open interest. | ||
| PRESENTER | Guanglian Hu, University of Sydney | |
| DISCUSSANT | Wenjun Zhang, Auckland University of Technology | |
| THURSDAY 15:00 to 15:30 | AFTERNOON TEA BREAK | WF8 Lounge |
| THURSDAY 15:30 to 17:30 | SESSION 3 | WF710 |
| CHAIRPERSON | Wenjun Zhang, Auckland University of Technology | |
| PAPER | INVESTOR SENTIMENT, UNEXPECTED INFLATION AND BITCOIN | |
| Thomas Conlon, University College Dublin | ||
| Shaen Corbet, Dublin City University & University of Waikato | ||
| Les Oxley, University of Waikato | ||
| Abstract: | ||
| The introduction of regulated CME futures contracts on Bitcoin in 2017 raised an expectation that cryptocurrencies would become part of mainstream financial markets. This also heightened links between traditional markets and Bitcoin, with the implication that the cryptocurrency would be subject to systematic spillovers. In this paper, we use high-frequency data to examine whether bitcoin basis risk is linked to investor sentiment from established financial markets. We present strong evidence that extreme investor sentiment, represented by volatility indices such as the VIX, is associated with a Bitcoin futures price that is lower than the spot. These findings are partially attributed to a coinciding increase in the relative volume of Bitcoin futures and have greater magnitude during periods of unexpected inflation and deflation. | ||
| PRESENTER | Les Oxley, University of Waikato | |
| DISCUSSANT | Yuan Lu, The Chinese University of Hong Kong | |
| PAPER | COMMODITY PREMIA AND RISK MANAGEMENT | |
| John Hua Fan, Griffith University | ||
| Tingxi Zhang Curtin University | ||
| Abstract: | ||
We examine the role of risk management in the context of commodity factor premia. Stopping losses in individual commodities effectively improves the average returns of long-short commodity premia through persistent reduction in the frequency and severity of drawdowns. The magnitude of improvement is related to the quality of the signal, commodity return volatility and autocorrelations, as well as transactions costs. The efficacy of a stop-loss strategy can be enhanced by dynamically calibrating loss thresholds in accordance with realized volatility, and it performs best in high conviction weighting schemes. Overall, we highlight the pivotal role of risk management beyond volatility targeting and risk-parity in harnessing commodity risk premia. | ||
| PRESENTER | Tingxi (Riven) Zhang, Curtin University | |
| DISCUSSANT | Les Oxley, University of Waikato | |
| PAPER | SKEWNESS, REALIZED VOLATILITY, AND OPTION PRICING | |
| Fang Liang, Sun Yat-sen University | ||
| Lingshan Du, Peking University, China | ||
| Abstract: | ||
| Efficiently exploiting information contained in price variations and accurately modelling the skewness of the underlying asset is critical for pricing options and other derivatives. In this paper, we propose a new and flexible option-pricing model that explicitly incorporates the dynamics of skewness and realized volatility. By the inverse Fourier transform, we derive closed-form option valuation formulas. Empirically, the model improves significantly upon benchmarks using S&P 500 index options. Overall, the joint modelling of skewness and realized volatility leads to an out-of-sample gain of 16.80% in pricing accuracy. The improvements are more pronounced for deep in-the-money calls, options with shorter maturities, and during highly volatile periods. | ||
| PRESENTER | Lingshan Du, Peking University | |
| DISCUSSANT | Wenqiang Liu, Auckland University of Technology | |
| PAPER | RESCALING THE MEAN-REVERTING 4/2 STOCHASTIC VOLATILITY MODEL FOR APPLICATIONS TO DERIVATIVE PRICING | |
| Jiling Cao, Auckland University of Technology | ||
| Jeong-Hoon Kim, Yonsei University | ||
| Wenjun Zhang, Auckland University of Technology | ||
| Wenqiang Liu, Auckland University of Technology | ||
| Abstract: | ||
| The 4/2 model, unifying the Heston and 3/2 models, exhibits important features of volatility and is somewhat tractable enough to provide a certain level of pricing procedure. However, a closed-form formula for derivative price is still lacking. We use a rescaling technique to obtain a closed-form formula for the approximate derivative price. Our formula has no integral term at all and it can be explicitly calculated by taking derivatives of the Black-Scholes price. Based on the analytic formula, we show that our model is more tractable than the original 4/2 model and yet flexible enough to capture important features of volatility. | ||
| PRESENTER | Wenqiang Liu, Auckland University of Technology | |
| DISCUSSANT | Lingshan Du, Peking University | |
| THURSDAY 18:30 to 21:30 | CONFERENCE DINNER | 4 SEASONS RESTAURANT |
END OF DAY ONE | ||
DAY 2: PROGRAMME OVERVIEW, FRIDAY 8TH SEPTEMBER 2023 | ||
|
FRIDAY 08:30 to 09:00 | ARRIVAL COFFEE AND TEA | WF8 Lounge |
| FRIDAY 09:00 to 10:00 | KEYNOTE ADDRESS |
STREAMED LIVE FROM USA IN ROOM WF710 |
| PROFESSOR GEERT ROUWENHORST, YALE UNIVERISTY | ||
Title: Why Contracts Fail Why is financial innovation so difficult? Using a novel comprehensive database of 230 surviving and defunct commodity futures contracts that traded on 28 exchanges between 1871 and 2022, we explore the factors that predict the probability of failure of a financial innovation following its introduction. The factors include the requirement of fair compensation for bearing risk, the incidence of extreme returns that challenge the fairness of contracts, competition across contracts and exchanges, and systemic shocks such as wars, economic recessions, and financial crises. Our results shed light on the conditions for the successful evolution of financial markets. | ||
|
FRIDAY 10:00 to 10:15 | MORNING TEA BREAK | WF8 Lounge |
| FRIDAY 10:15 to 12:15 | SESSION 4 | WF710 |
| CHAIRPERSON | Raymond Kim, Northern Arizona University | |
| PAPER | TRADING COSTS AND MARKET MICROSTRUCTURE INVARIANCE: IDENTIFYING BET ACTIVITY | |
| Ai Jun Hou, Stockholm University | ||
| Lars L. Norden, Stockholm University | ||
| Caihong Xu, Stockholm University | ||
| Abstract: | ||
| Market microstructure invariance (MMI) stipulates that trading costs of financial assets are driven by the volume and volatility of bets, but these variables are inherently difficult to identify. With futures transactions data, we estimate bet volume as the trading volume of brokerage firms that trade on behalf of their clients and bet volatility as the trade-related component of futures volatility. We find that the futures bid-ask spread lines up with bet volume and bet volatility as predicted by MMI, and that intermediation by high-frequency traders does not interfere with the MMI relation. | ||
| PRESENTER | Ai Jun Hou, Stockholm University | |
| DISCUSSANT | Ion Lucas Saru, VU Amsterdam and Tinbergen Institute | |
| PAPER | WHO KNOWS? INFORMATION DIFFERENCES BETWEEN TRADER TYPES | |
| Albert J. Menkveld, VU Amsterdam and Tinbergen Institute | ||
| Ion Lucas Saru, VU Amsterdam and Tinbergen Institute | ||
| Abstract: | ||
| We study the informativeness of agent and principal trades. Order informativeness depends on the horizon and frequency we analyse. In line with the literature on high-frequency trading, principals are more informed than agents at the highest frequency, as measured by the contribution of the respective order flow series to the variance of efficient price innovations. Once we move to lower frequencies, price discovery is dominated by agents, while the share of principals goes to zero. This is reflected in the gross trading revenues of agents and principals at different frequencies. Our results hold across market conditions as measured by the VIX. | ||
| PRESENTER | Ion Lucas Saru, VU Amsterdam and Tinbergen Institute | |
| DISCUSSANT | Raymond Kim, Northern Arizona University | |
| PAPER | INTEREST RATE HEDGING AND SILICON VALLEY BANK IDIOSYNCRASIES | |
| Raymond Kim, Northern Arizona University | ||
| Abstract: | ||
| Under a "mea-culpa" framework, evidence suggests that financial institutions practice discretionary hedging of both interest rate and funding risks, unlike Silicon Valley Bank and First Republic Bank. Banks asymmetrically manage risk by intensifying hedging with interest rate derivatives as HTM and AFS portfolio losses accrue, and by reducing hedging intensity as portfolio gains accrue. As funding risk increases, banks also intensify hedging, suggesting the mistakes of Silicon Valley Bank are idiosyncratic, not systematic. Evidence suggests financial institutions are generally successful at incorporating forward interest rate guidance when managing and anticipating balance sheet risks. | ||
| PRESENTER | Raymond Kim, Northern Arizona University | |
| DISCUSSANT | Ni Yang, Auckland University of Technology | |
| PAPER | DELTA FLUCTUATIONS AND OPTION RETURNS | |
| Yuan Lu, The Chinese University of Hong Kong | ||
| Abstract: | ||
| The paper documents a significant, robust positive relationship between delta fluctuations and option returns. As absolute delta fluctuations introduce equal risks to both option buyers and sellers, the return predictability of delta fluctuations cannot be attributed to a rational risk-based explanation. Instead, it stems from the asymmetrical risk perceptions of option buyers and sellers. Our findings suggest that option buyers play a dominant role in pricing the delta fluctuations, while the option sellers are inclined to be more “risk-seeking”. The sellers’ relatively lower awareness of risk contributes to the mispricing of delta fluctuations in options. Furthermore, we explore the time-series and cross-sectional variations of this option mispricing and find that, when limits to arbitrage are higher, the return predictability of delta fluctuations becomes more prominent. | ||
| PRESENTER | Yuan Lu, The Chinese University of Hong Kong | |
| DISCUSSANT | Ai Jun Hou, Stockholm University | |
| FRIDAY 12:15 to 12:30 | CLOSING ADDRESS and AWARDSPaper Award and Closing Remarks | WF710 |
| FRIDAY 12:30 TO 13:30 | LUNCH | WF8 Lounge |