Francesco Audrino

Realized Correlation Tick-By-Tick

We propose the Heterogeneous Autoregressive (HAR) model for the estimation and prediction of realized correlations. We construct a realized correlation measure where both the volatilities and the covariances are computed from tick-by-tick data. As for the realized volatility, the presence of market microstructure can induce significant bias in standard realized covariance measure computed with artificially regularly spaced returns. Contrary to these standard approaches we analyse a simple and unbiased realized covariance estimator that does not resort to the construction of a regular grid, but directly and efficiently employs the raw tick-by-tick returns of the two series. Montecarlo simulations calibrated on realistic market microstructure conditions show that this simple tick-by-tick covariance possesses no bias and the smallest dispersion among the covariance estimators considered in the study. In an empirical analysis on S&P 500 and US bond data we find that realized correlations show significant regime changes in reaction to financial crises. Such regimes must be taken into account to get reliable estimates and forecasts.

Tree‐structured generalized autoregressive conditional heteroscedastic models

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 fitting of such trees is constructed within the likelihood framework for non-Gaussian observations: it is very different from the well-known regression tree procedure which is based on residual sums of squares. Our strategy includes the classical GARCH model as a special case and allows us to increase model complexity in a systematic and flexible way. We derive a consistency result and conclude from simulation and real data analysis that the new method has better predictive potential than other approaches.

Lassoing the HAR Model: A Model Selection Perspective on Realized Volatility Dynamics

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 out-of-sample performance. We prove that the least absolute shrinkage and selection operator (Lasso) recovers the lags structure of the HAR model asymptotically if it is the true model, and we present Monte Carlo evidence in finite samples. The HAR models lags structure is not fully in agreement with the one found using the Lasso on real data. Moreover, we provide empirical evidence that there are two clear breaks in structure for most of the assets we consider. These results bring into question the appropriateness of the HAR model for realized volatility. Finally, in an out-of-sample analysis, we show equal performance of the HAR model and the Lasso approach.

A General Multivariate Threshold GARCH Model With Dynamic Conditional Correlations

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 study the empirical performance of our model in two applications using U.S. stock and bond market data. In both applications our model has, in terms of statistical and economic significance, higher forecasting power than several other multivariate GARCH models for conditional correlations.

The impact of general non-parametric volatility functions in multivariate GARCH models

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 the product of conditional correlations and individual volatilities, it is plausible that improving the estimation of individual volatilities will lead to better covariance forecasts, too. Functional gradient descent (FGD) has already been shown to improve substantially in-sample and out-of-sample covariance accuracy in the very simple constant conditional correlation (CCC) setting. Following this direction, the impact of FGD volatility estimates is tested in several multivariate GARCH settings, both at the multivariate and at the univariate portfolio levels. In particular, improving conditional correlations and improving individual volatilities are compared, to establish which effect produces the best fits and predictions for conditional covariances.

An Empirical Analysis of the Ross Recovery Theorem

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 investigate the predictive information content of its moments. We find that changes in the recovered moments can be used to time the index, yielding strategies that not only outperform the market, but are also significantly less volatile.

A dynamic model of expected bond returns: A functional gradient descent approach

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 outperforms the classical univariate analysis used in the literature.

Semi-parametric forecasts of the implied volatility surface using regression trees

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. To overcome the poor predictive power of existing models, we include a grid in the region of interest, and implement a cross-validation strategy to find an optimal stopping value for the boosting procedure. Back testing the out-of-sample performance on a large data set of implied volatilities from S&P 500 options, we provide empirical evidence of the strong predictive power of our model.

Volatility Forecasting: Downside Risk, Jumps and Leverage Effect

We provide empirical evidence of 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 quadratic variation. The leverage effect is separated into continuous and discontinuous effects, and past volatility is separated into “good” and “bad”, as well as into continuous and discontinuous risks. Using a long history of the S & P500 price index, we find that the continuous leverage effect lasts about one week, while the discontinuous leverage effect disappears after one day. “Good” and “bad” continuous risks both characterize the volatility persistence, while “bad” jump risk is much more informative than “good” jump risk in forecasting future volatility. The volatility forecasting model proposed is able to capture many empirical stylized facts while still remaining parsimonious in terms of the number of parameters to be estimated.

Tree-Structured Multiple Regimes in Interest Rates

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 generalized autoregressive conditional heteroscedasticity and square root processes as simple special cases. It allows us to estimate the optimal number of regimes endogenously from the data and to exploit possible additional information in the term structure and in other macroeconomic variables. We provide empirical evidence of the strong potential of the GTS model in forecasting conditional first and second moments, also in comparison with alternative models of the short rate.