Francesco Audrino

University of St.Gallen · Department of Economics

We provide empirical evidence on volatility forecasting in relation to asymmetries present in the dynamics of both return and volatility processes. Using recently developed methodologies to detect jumps from high frequency price data, we estimate the size of positive and negative jumps and propose a methodology to estimate the size of jumps in the...

We propose a new realized volatility model which specifically accounts for memory-and leverage-effect through novel price pattern recognition based on an open-minded learning environment in convolutional neural networks (CNN), extending the heterogeneous autoregressive (HAR) model. More specifically, we investigate for the first time the inclusion...

This paper builds upon previous research findings that show macro sentiment data-augmented models are better at predicting the yield curve. We extend the dynamic Nelson-Siegel model with macro sentiment data from either Twitter or RavenPack. Vector autogressive (VAR) models and Markov-switching VAR models are used to predict changes in the shape of...

We investigate whether Lasso-type linear methods are able to improve the predictive accuracy of OLS in selecting relevant firm characteristics for forecasting the future cross-section of stock returns. Through extensive Monte Carlo simulations we show that Lasso-type predictions are superior to OLS when type II errors are a concern. The results cha...

We revisit the role played by sentiment extracted from news articles related to earnings announcements as a driver of firms' return, volatility, and trade volume dynamics. To this end we apply causal machine learning on the earnings announcements of a wide cross-section of US companies. This approach allows us to investigate firms' price and volume...

We empirically investigate how retail and institutional investor attention is related to the way stock markets process information. With a focus on 360 US stocks in the S&P 500 universe, our results show that higher retail investors’ attention around news releases increases the post-announcement stock return volatility, whereas institutional invest...

Building on the method of Ludwig (2015) to construct robust state price density surfaces from snapshots of option prices, we develop a nonparametric estimation strategy based on the recovery theorem of Ross (2015). Using options on the S&P 500, we then investigate whether or not recovery yields predictive information beyond what can be gleaned from...

We revisit the role played by sentiment extracted from news articles related to earnings announcements as a driver of firms' return, volatility, and trade volume dynamics. To this end we apply causal machine learning on the earnings announcements of a wide cross-section of US companies. This approach allows us to investigate firms' price and volume...

We empirically investigate how retail and institutional investor attention is related to the way stock markets process information. With a focus on 360 US stocks in the S&P 500 universe, our results show that higher retail investors' attention around news releases increases the post-announcement stock return volatility, whereas institutional invest...

Are the financially and institutionally strongest clubs capable of systematically reaching the top positions in the European national football leagues treated differently in terms of awarded sanctions because of the external off the pitch pressure they can put on match officials? This study helps shed some light on this controversial question fierc...

We analyze the impact of sentiment and attention variables on stock market volatility by using a novel and extensive dataset that combines social media, news articles, information consumption, and search engine data. Applying a state-of-the-art sentiment classication technique, we investigate the question of whether sentiment and attention measures...

The Heterogeneous Autoregressive (HAR) model is commonly used in modeling the dynamics of realized volatility. In this paper, we propose a flexible HAR(1,. .. , p) specification , employing the adaptive LASSO and its statistical inference theory to see whether the lag structure (1, 5, 22) implied from an economic point of view can be recovered by s...

The fast-growing literature on news analytics provides evidence that financial markets are partially driven by sentiments. In contrast with previous studies that have almost exclusively focused on the direct effects of the news related to single companies or sectors, we investigate the time-varying dynamics of news' cross-industry influences for a...

We introduce a wild multiplicative bootstrap for M and GMM estimators in nonlinear models when autocorrelation structures of moment functions are unknown. The implementation of the bootstrap algorithm does not require any parametric assumptions on the data generating process. After proving its validity, we also investigate the accuracy of our proce...

Scientia, the leading science communication publication, just published an outreach article describing one of my ongoing research projects funded by the Swiss National Science Foundation. In the SentiVol project, together with my collaborators Fabio Sigrist from the Lucerne University of Applied Sciences and Arts and Daniele Ballinari from the Uni...

Are the strongest clubs capable of reaching the top positions systematically in the Eu-ropean national football leagues treated dierently by match ocials in terms of awarded sanctions? This study helps shed some light on this controversial question ercely debated among fans and sports journalists. Analyzing data on the top ve European leagues for t...

We address the fiercely debated question of whether the strongest European football clubs get special, preferential treatment from match officials in their decisions on the teams' players over the course of the teams' trophy winning streaks. To give an empirical answer to this question, we apply a rigorous econometric analysis for causal effect est...

We propose a new approach based on a generalization of the logit model to improve prediction accuracy in US bank failures. Mixed-data sampling (MIDAS) is introduced in the context of a logistic regression. We also mitigate the class-imbalance problem in data and adjust the classification accuracy evaluation. In applying the suggested model to the p...

We analyze the impact of sentiment and attention variables on volatility by using a novel and extensive dataset that combines social media, news articles, information consumption, and search engine data. Applying a state-of-the-art sentiment classification technique, we investigate the question of whether sentiment and attention measures contain ad...

We derive new theoretical results on the properties of the adaptive least absolute shrinkage and selection operator (adaptive lasso) for possibly nonlinear time series models. In particular, we investigate the question of how to conduct inference on the parameters given an adaptive lasso model. Central in this study is the test of the hypothesis th...

A (conservative) test is applied to investigate the optimal lag structure for modeling realized volatility dynamics. The testing procedure relies on the recent theoretical results that show the ability of the adaptive least absolute shrinkage and selection operator (adaptive lasso) to combine efficient parameter estimation, variable selection, and...

A (conservative) test is applied to investigate the optimal lag structure for modelingrealized volatility dynamics. The testing procedure relies on the recent theoretical results that showthe ability of the adaptive least absolute shrinkage and selection operator (adaptive lasso) to combinee cient parameter estimation, variable selection, and valid...

The Heterogeneous Autoregressive (HAR) model is commonly used in modeling the dynamics of realized volatility. In this paper, we propose a flexible HAR(1,...,p) specification, employing the adaptive LASSO and its statistical inference theory to see whether the lag structure (1, 5, 22) implied from an economic point of view can be recovered by stati...

We introduce a wild multiplicative bootstrap for M and GMM estimators in nonlinear models when autocorrelation structures of moment functions are unknown. The implementation of the bootstrap algorithm does not require any parametric assumptions on the data generating process. After proving its validity, we also investigate the accuracy of our proce...

Realized volatility computed from high-frequency data is an important measure for many applications in finance, and its dynamics have been widely investigated. Recent notable advances that perform well include the heterogeneous autoregressive (HAR) model which can approximate long memory, is very parsimonious, is easy to estimate, and features good...

As a means of validating an option pricing model, we compare the ex-post intra-day realized variance of options with the realized variance of the associated underlying asset that would be implied using assumptions as in the Black and Scholes (BS) model, the Heston and the Bates model. Based on data for the S&P 500 index, we find that the BS model i...

A (conservative) test is constructed to investigate the optimal lag structure for forecasting realized volatility dynamics. The testing procedure relies on the recent theoretical results that show the ability of the adaptive least absolute shrinkage and selection operator (adaptive lasso) to combine e�cient parameter estimation, variable selection,...

We develop a multivariate dynamic term structure model, which takes into account the nonlinear (time-varying) relation between interest rates and the state of the economy. In contrast to the classical term structure literature, in which nonlinearities are captured by increasing the number of latent state variables or by latent regime shifts, in our...

The predictive power of recently introduced components affecting correlations is investigated. The focus is on models allowing for a flexible specification of the short-run component of correlations as well as the long-run component. Moreover, models allowing the correlation dynamics to be subjected to regime-shift caused by threshold-based structu...

Building on the results of Ludwig (2012), we propose a method to construct robust time-homogeneous Markov chains that capture the risk-neutral transition of state prices from current snapshots of option prices on the S&P 500 index. Using the recovery theorem of Ross (2013), we then derive the market’s forecast of the real-world return density and i...

We derive new theoretical results on the properties of the adaptive least absolute shrinkage and selection operator (adaptive lasso) for time series regression models. In particular, we investigate the question of how to conduct finite sample inference on the parameters given an adaptive lasso model for some fixed value of the shrinkage parameter....

Realized volatility computed from high-frequency data is an important measure for many applications in finance. However, its dynamics are not well understood to date. Recent notable advances that perform well include the heterogeneous autoregressive (HAR) model which is economically interpretable, easy to estimate, and features good out-of-sample p...

We suggest a joint analysis of ex-post intra-day variability in an option and its associated underlying asset market as a means of validating an option pricing model. For this purpose, we contrast option realized variance with the realized variance that would be implied from the underlying asset price path under certain model assumptions. In the em...

We develop a multivariate dynamic term structure model, which takes into account the nonlinear (time-varying) relationship between interest rates and the state of the economy. In contrast to the classical term structure literature, where nonlinearities are captured by increasing the number of latent state variables, or by latent regime shifts, in o...

We propose a new methodology to estimate the empirical pricing kernel implied from option data. In contrast to most of the studies in the literature that use an indirect approach, i.e. first estimating the physical and risk-neutral densities and obtaining the pricing kernel in a second step, we follow a direct approach. Departing from an adequate p...

This paper presents two classes of tick-by-tick covariance estimators adapted to the case of rounding in the price time stamps to a frequency lower than the typical arrival rate of tick prices. Through Monte Carlo simulations we investigate the behavior of such estimators under realistic market microstructure conditions analogous to those of the fi...

IntroductionStylized Facts on Realized VolatilityHeterogeneity and Volatility PersistenceHAR ExtensionsMultivariate ModelsApplicationsConclusion

Functional gradient descent (FGD), a recent technique coming from computational statistics, is applied to the estimation of the conditional moments of the short rate process with the goal of finding the main drivers of the drift and volatility dynamics. FGD can improve the accuracy of some reasonable starting estimates obtained using classical shor...

Motivated by the need for an unbiased and positive-semidefinite estimator of multivariate realized covariance matrices, we model noisy and asynchronous ultra-high-frequency asset prices in a state-space framework with missing data. We then estimate the covariance matrix of the latent states through a Kalman smoother and Expectation Maximization (KE...

We investigate whether a more sophisticated technique able to forecast accurately the future movements of the implied volatility surface may help in improving the performance of basic option strategies. To this goal we construct a set of strategies using predicted option returns for a forecasting period of ten trading days and form profitable hold-...

In this paper we propose a smooth transition tree model for both the conditional mean and variance of the short-term interest rate process. The estimation of such models is addressed and the asymptotic properties of the quasi-maximum likelihood estimator are derived. Model specification is also discussed. When the model is applied to the US short-t...

We provide new empirical evidence on volatility forecasting in relation to asymmetries present in the dynamics of both return and volatility processes. Leverage and volatility feedback effects of the S&P 500 price and volatility dynamics are examined using recently developed methodologies to detect jumps and to disentangle their size from continuou...

Revised version of paper no. 2005-04. We propose a new multivariate GARCH model with Dynamic Conditional Correlations that extends previous models by admitting multivariate thresholds in conditional volatilities and correlations. The model estimation is feasible in large dimensions and the positive deniteness of the conditional covariance matrix is...

We present a new semi-parametric model for the prediction of implied volatility surfaces that can be estimated using machine learning algorithms. Given a reasonable starting model, a boosting algorithm based on regression trees sequentially minimizes generalized residuals computed as differences between observed and estimated implied volatilities....

We propose a simple but effective estimation procedure to extract the level and the volatility dynamics of a latent macroeconomic factor from a panel of observable indicators. Our approach is based on a multivariate conditionally heteroskedastic exact factor model that can take into account the heteroskedasticity feature shown by most macroeconomic...

We propose constructing a set of trading strategies using predicted option returns for a relatively small forecasting period of ten trading days to form profitable hold-to-expiration, equally weighted, zero-cost portfolios based on 1-month at-the-money call and put options. We use a statistical machine learning procedure based on regression trees t...

In this paper we propose a smooth transition tree model for both the conditional mean and the conditional variance of the short-term interest rate process. Our model incorporates the interpretability of regression trees and the flexibility of smooth transition models to describe regime switches in the short-term interest rate series. The estimation...

http://ideas.repec.org/p/usg/dp2007/2007-42.html

A tree-structured heterogeneous autoregressive (tree-HAR) process is proposed as a simple and parsimonious model for the estimation and prediction of tick-by-tick realized correlations. The model can account for different time and other relevant predictors' dependent regime shifts in the conditional mean dynamics of the realized correlation series....

This paper presents two classes of tick-by-tick covariance estimators adapted to the case of rounding in the price time stamps to a frequency lower than the typical arrival rate of tick prices. Through Monte Carlo simulations, we investigate the behavior of such estimators under realistic market microstructure conditions analogous to those of the f...

http://ideas.repec.org/p/usg/dp2008/2008-04.html

We propose a new semi-parametric model for the implied volatility surface, which incorporates machine learning algorithms. Given a starting model, a tree-boosting algorithm sequentially minimizes the residuals of observed and estimated implied volatility. To overcome the poor predicting power of existing models, we include a grid in the region of i...

We propose a tick-by-tick covariance estimator adapted to the case of rounding in the price time stamps and investigate, through Monte Carlo simulations, the behavior of such estimator under market microstructure conditions analogous to that of the financial data studied in the empirical part. Moreover, we provide empirical evidence that the Hetero...

We propose a multivariate nonparametric technique for generating reliable short-term historical yield curve scenarios and confidence intervals. The approach is based on a Functional Gradient Descent (FGD) estimation of the conditional mean vector and covariance matrix of a multivariate interest rate series. It is computationally feasible in large d...

We propose a flexible GARCH-type model for the prediction of volatility in financial time series. The approach relies on the idea of using multivariate B-splines of lagged observations and volatilities. Estimation of such a B-spline basis expansion is constructed within the likelihood framework for non-Gaussian observations. As the dimension of the...

We apply the recently introduced generalized tree-structured (GTS) model to the analysis and forecast of stock market diversity. Diversity is a measure of capital concentration across a market that plays a central role in the search for arbitrage. The GTS model allows for different conditional mean and volatility regimes that are directly related t...

The paper examines the performance of four multivariate volatility models, namely CCC, VARMA-GARCH, DCC and BEKK, for the crude oil spot and futures returns of two major benchmark international crude oil markets, Brent and WTI, to calculate optimal portfolio weights and optimal hedge ratios, and to suggest a crude oil hedge strategy. The empirical...

We consider the estimation of integrated covariance matrices of high dimensional diffusion processes by using high frequency data. We start by studying the most commonly used estimator, the realized covariance matrix (RCV). We show that in the high dimensional case when the dimension p and the observation frequency n grow in the same rate, the limi...

A multivariate methodology based on functional gradient descent to estimate and forecast time-varying expected bond returns is presented and discussed. Backtesting this procedure on US monthly data, empirical evidence of its strong forecasting potential in terms of the accuracy of the predictions is collected. The proposed methodology clearly outpe...

This article develops a generalized tree-structured (GTS) model of the short-term interest rate that accommodates regime-dependent mean reversion and regime-dependent volatility clustering and level effects in the conditional variance. The model is constructed using the idea of multivariate tree-structured thresholds and nests the popular generaliz...

Recent studies have revealed that financial volatilities and correlations move together over time across assets and markets. The main effort has been on improving the flexibility of conditional correlation dynamics, while maintaining computational feasibility for large estimation problems. However, since in such models conditional covariances are t...

We propose a simple class of multivariate GARCH models, allowing for time-varying conditional correlations. Estimates for time-varying conditional correlations are constructed by means of a convex combination of averaged correlations (across all series) and dynamic realized (historical) correlations. Our model is very parsimonious. Estimation is co...

We propose an affine term structure model which accommodates nonlinearities in the drift and volatility function of the short-term interest rate. Such nonlinearities are a consequence of discrete beta-distributed regime shifts constructed on multiple thresholds. We derive iterative closed-form formula for the whole yield curve dynamics that can be...

We propose a non-parametric local likelihood estimator for the log-transformed autoregressive conditional heteroscedastic (ARCH) (1) model. Our non-parametric estimator is constructed within the likelihood framework for non-Gaussian observations: it is different from standard kernel regression smoothing, where the innovations are assumed to be norm...

The estimation and forecast of the volatility matrix are two of the main tasks of financial econometrics since they are essential ingredients in many practical applications. Unfortunately the use of classical multivariate methods in large dimensions is difficult because of the curse of dimensionality. We present a general semiparametric technique,...

It is difficult to compute Value-at-Risk (VaR) using multivariate models able to take into account the dependence structure between large numbers of assets and being still computationally feasible. A possible procedure is based on functional gradient descent (FGD) estimation for the volatility matrix in connection with asset historical simulation....

We propose a non-parametric local likelihood estimator for the log-transformed autoregressive conditional heteroscedastic (ARCH) (1) model. Our non-parametric estimator is constructed within the likelihood framework for non-Gaussian observations: it is different from standard kernel regression smoothing, where the innovations are assumed to be norm...

The daily term structure of interest rates is filtered to reduce the influence of cross-correlations and autocorrelations on its factors. A three-factor model is fitted to the filtered data. We perform statistical tests, finding that factor loadings are unstable through time for daily data. This finding is not due to the presence of outliers nor to...

A tree-structured heterogeneous autoregressive (tree-HAR) process is proposed as a simple and parsimonious model for the estimation and prediction of tick-by-tick realized correlations. The model can account for different time and other relevant predictors' dependent regime shifts in the conditional mean dynamics of the realized correlation series....

Matching university places to students is not as clear cut or as straightforward as it ought to be. By investigating the matching algorithm used by the German central clearinghouse for university admissions in medicine and related subjects, we show that a procedure designed to give an advantage to students with excellent school grades actually harm...

We introduce a new multivariate GARCH model with multivariate thresholds in conditional correlations and develop a two-step estimation procedure that is feasible in large dimensional applications. Optimal threshold functions are estimated endogenously from the data, and the model conditional covariance matrix is ensured to be positive definite. We...

We propose a general double tree structured AR-GARCH model for the analysis of global equity index returns. The model extends previous approaches by incorporating (i) several multivariate thresholds in conditional means and volatilities of index returns and (ii) a richer specification for the impact of lagged foreign (US) index returns in each thre...

We propose a functional gradient descent algorithm (FGD) for estimating volatility and conditional covariances (given the past) for very high-dimensional financial time series of asset price returns. FGD is a kind of hybrid of nonparametric statistical function estimation and numerical optimization. Our FGD algorithm is computationally feasible in...

Prices or returns of financial assets are most often collected in local times of the trading markets. The need to synchronize multivariate time series of financial prices or returns is motivated by the fact that information continues to flow for closed markets while others are still open. We propose here a synchronization technique which takes this in...

We propose a new generalized autoregressive conditional heteroscedastic (GARCH) model with tree-structured multiple thresholds for the estimation of volatility in financial time series. The approach relies on the idea of a binary tree where every terminal node parameterizes a (local) GARCH model for a partition cell of the predictor space. The fitt...

We propose an empirical approach to determine the various economic sources driving the US yield curve. We allow the conditional dynamics of the yield at different maturities to change in reaction to past information coming from several relevant predictor variables. We consider both endogenous, yield curve factors and exogenous, macroeconomic factor...

Diss. no. 14565 math. SFIT Zurich. Literaturverz.

http://ideas.repec.org/p/usg/dp2005/2005-04.html