Publications

On this part of the site you can find a list of my publications, with link to the journal version and the most recent working paper version. For questions, suggestions, comments, or anything else related to my research please feel free to contact me at: vanderwel@ese.eur.nl. For a list without abstracts, click here.

Modelling Sovereign Credit Ratings: Evaluating the Accuracy and Driving Factors using Machine Learning Techniques. Computational Economics (2022)
Joint work with Bart Overes
Abstract: Sovereign credit ratings summarize the creditworthiness of countries. These ratings have a large influence on the economy and the yields at which governments can issue new debt. This paper investigates the use of a Multilayer Perceptron (MLP), Classification and Regression Trees (CART), Support Vector Machines (SVM), Naïve Bayes (NB), and an Ordered Logit (OL) model for the prediction of sovereign credit ratings. We show that MLP is best suited for predicting sovereign credit ratings, with a random cross-validated accuracy of 68%, followed by CART (59%), SVM (41%), NB (38%), and OL (33%). Investigation of the determining factors shows that there is some heterogeneity in the important variables across the models. However, the two models with the highest out-of-sample predictive accuracy, MLP and CART, show a lot of similarities in the influential variables, with regulatory quality, and GDP per capita as common important variables. Consistent with economic theory, a higher regulatory quality and/or GDP per capita are associated with a higher credit rating.
Please click on the title to find the paper on Springer (it is open access), or click here for a late working paper version.

An Asset Pricing Approach to Testing General Term Structure Models. Journal of Financial Economics (2019) 134:1, p165-191
Joint work with Bent Jesper Christensen
Abstract: We develop a new empirical approach to term structure analysis that allows testing for time-varying risk premia and arbitrage opportunities in models with both unobservable factors and factors identified as the innovations to observed macroeconomic variables. Factors may play double roles as both covariance-generating common shocks driving yields and determinants of market prices of risk in cross-sectional pricing. The evidence favors time-varying risk prices significantly related to the second Stock-Watson principal component of macroeconomic variables and to changes in the industrial production index. Our preferred specification includes these two observable and two unobservable factors, with the no-arbitrage condition imposed.
Please click on the title to find the paper on Science Direct website, or click here for a late working paper version.

What do Professional Forecasters actually predict? International Journal of Forecasting (2018) 34:2, p288-311
Joint work with Didier Nibbering and Richard Paap
Abstract: In this paper we study what professional forecasters predict. We use spectral analysis and state space modeling to decompose economic time series into a trend, a business-cycle, and irregular component. To examine which components are captured by professional forecasters, we regress their forecasts on the estimated components extracted from both the spectral analysis and the state space model. For both decomposition methods we find that the Survey of Professional Forecasters in the short run can predict almost all variation in the time series due to the trend and the business-cycle, but the forecasts contain little or no significant information about the variation in the irregular component.
Please click on the title to find the paper on Science Direct website, or click here for a late working paper version.

Combining Density Forecasts Using Focused Scoring Rules. Journal of Applied Econometrics (2017) 32:7, p1298-1313
Joint work with Anne Opschoor and Dick van Dijk
Abstract: We investigate the added value of combining density forecasts focused on a specific region of support. We develop forecast combination schemes that assign weights to individual predictive densities based on the censored likelihood scoring rule and the continuous ranked probability scoring rule (CRPS) and compare these to weighting schemes based on the log score and the equally weighted scheme. We apply this approach in the context of measuring downside risk in equity markets using recently developed volatility models, including HEAVY, realized GARCH and GAS models, applied to daily returns on the S&P 500, DJIA, FTSE and Nikkei indexes from 2000 until 2013. The results show that combined density forecasts based on optimizing the censored likelihood scoring rule significantly outperform pooling based on equal weights, optimizing the CRPS or log scoring rule. In addition, 99% Value-at-Risk estimates improve when weights are based on the censored likelihood scoring rule.
Please click on the title to find the paper on the Wiley website [the article is open access!], or click here for a late working paper version. The data are available through the JAE website.

Intraday Price Discovery in Fragmented Markets. Journal of Financial Markets (2017) 32, p28-48
Joint work with Sait Ozturk and Dick van Dijk
Abstract: We explore intraday variation in the contribution to price discovery across different exchanges. We estimate a structural model with time-varying parameters in state space form using Maximum Likelihood. We analyze data for 50 S&P 500 stocks in 2013 and find that the constancy of shares in price discovery is rejected. Tighter quoted spreads attract informed trading from other exchanges. Exchange listing and industrial sector of a stock significantly affect the dominant venues of price discovery in different parts of the day and following macroeconomic news announcements.
Please click on the title to find the paper on the Science Direct website, or click here for a late working paper version. The web appendix is available with a link on the first page of the working paper version.

Estimating Dynamic Equilibrium Models using Mixed Frequency Macro and Financial Data. Journal of Econometrics (2016) 194:1, p116-137
Joint work with Bent Jesper Christensen and Olaf Posch
Abstract: We provide a framework for inference in dynamic equilibrium models including financial market data at daily frequency, along with macro series at standard lower frequency. Our formulation of the macro-finance model in continuous-time conveniently accounts for the difference in observation frequency. We suggest the use of martingale estimating functions (MEF) to infer the structural parameters of the model directly through a nonlinear optimization scheme. This method is compared to regression-based methods and the general method of moments (GMM). We illustrate our approaches by estimating various versions of the AK-Vasicek model with mean-reverting interest rates. We provide asymptotic theory and Monte Carlo evidence on the small sample behavior of the estimators and report empirical estimates using 30 years of U.S. macro and financial data.
Please click on the title to find the paper on the Science Direct website, or click here for a late working paper version. The web appendix is included at the end of the working paper version.

Market Set-Up in Advance of Federal Reserve Policy Rate Decisions. Economic Journal (2016) 126, p618-653
Joint work with Dick van Dijk and Robin Lumsdaine
Abstract: This article considers the extent to which the federal (fed) funds futures market prepares for Federal Open Market Committee (FOMC) announcements. We demonstrate that there is often less variation in fed funds futures prices during the period immediately preceding an FOMC announcement than in earlier periods, despite greater trading activity, as the market has already incorporated anticipated signals. We find that macro announcements and central bank officials' congressional testimony are of comparable importance, whereas speeches are relatively unimportant. In addition, macro announcements have stronger effects when they are released during the Fed's ‘blackout’ period, emphasising important interaction effects.
Please click on the title to find the paper on the Wiley website, or click here for a late working paper version. Supporting information (the web appendix and a zip with all code) is available also from the Wiley website.

Dynamic Factor Models for the Volatility Surface. Advances in Econometrics (2016) 35, p127-174
First author, joint work with Sait Ozturk and Dick van Dijk
Abstract: The implied volatility surface is the collection of volatilities implied by option contracts for different strike prices and time-to-maturity. We study factor models to capture the dynamics of this three-dimensional implied volatility surface. Three model types are considered to examine desirable features for representing the surface and its dynamics: a general dynamic factor model, restricted factor models designed to capture the key features of the surface along the moneyness and maturity dimensions, and in-between spline-based methods. Key findings are that: (i) the restricted and spline-based models are both rejected against the general dynamic factor model, (ii) the factors driving the surface are highly persistent, (iii) for the restricted models option Delta is preferred over the more often used strike relative to spot price as measure for moneyness.
Please click on the title to find the paper on the Emerald website, or click here for a late working paper version. The appendix tables and figures are included at the end of both the journal and working paper version.

Forecasting Interest Rates with Shifting Endpoints. Journal of Applied Econometrics (2014) 29:5, p693-712
Joint work with Dick van Dijk, Siem Jan Koopman and Jonathan Wright
Abstract: Existing studies on interest rate forecasting either treat yields as being stationary around a fixed mean or as a random walk process. In this study we consider forecasting the term structure of interest rates with the assumption that the yield curve is driven by factors that are stationary around a slowly time-varying mean or "shifting endpoint". The shifting endpoints are captured using either (i) time series methods (exponential smoothing), or (ii) long-range survey forecasts of either interest rates or inflation and output growth, or (iii) exponentially smoothed realizations of these macro variables. We find that allowing for shifting endpoints in yield curve factors can provide gains in the out-of-sample predictive accuracy, relative to stationary and random walk benchmarks. These gains are statistically significant, and can involve more than 20 percent reductions in root mean square prediction error.
Please click on the title to find the paper on the Wiley website, or click here for a late working paper version. The SSRN version includes the web-appendix at the back of the PDF. The data are available through the JAE website.

Smooth Dynamic Factor Analysis with Application to the U.S. Term Structure of Interest Rates. Journal of Applied Econometrics (2014) 29:1, p65-90
Joint work with Borus Jungbacker and Siem Jan Koopman
Abstract: We consider the dynamic factor model and show how smoothness restrictions can be imposed on factor loadings by using cubic spline functions. We develop statistical procedures based on Wald, Lagrange multiplier and likelihood ratio tests for this purpose. The methodology is illustrated by analyzing a newly updated monthly time series panel of U.S. term structure of interest rates. Dynamic factor models with and without smooth loadings are compared with dynamic models based on Nelson-Siegel and cubic spline yield curves. We conclude that smoothness restrictions on factor loadings are supported by the interest rate data and can lead to more accurate forecasts.
Please click on the title to find the paper on the Wiley website, or click here for a late working paper version. The data are available through the JAE website. The first working paper version from March 2009 contains some additional results, including in-sample tests for affine models and alternative factor dynamic specifications.

Predicting Volatility and Correlations with Financial Conditions Indexes. Journal of Empirical Finance, (2014) 29, p435-447
Joint work with Anne Opschoor and Dick van Dijk
Abstract: We model the impact of financial conditions on asset market volatilities and correlations. We extend the Spline-GARCH model for volatility and DCC model for correlation to allow for inclusion of indexes that measure financial conditions. In our empirical application we consider daily stock returns of US deposit banks during the period 1994-2011, and proxy financial conditions by the Bloomberg Financial Conditions Index (FCI) which comprises the money, bond, and equity markets. We find that worse financial conditions are associated with both higher volatility and higher correlations between stock returns, especially during crises. Moreover, including the FCI in volatility and correlation modeling improves Value-at-Risk estimates, particularly at short horizons.
Please click on the title to find the paper on the Science Direct website, or click here for a late working paper version.

Order Flow and Volatility: An Empirical Investigation. Journal of Empirical Finance (2014) 28, p185-201
Joint work with Anne Opschoor, Dick van Dijk and Nick Taylor
Abstract: We study the relationship between order flow and volatility. To this end we develop a comprehensive framework that simultaneously controls for the effects of macro announcements and order flow on prices and the effect of macro announcements on volatility. Using high-frequency 30-year U.S. Treasury bond futures data, we find a statistically and economically significant relationship between the absolute value of order flow and volatility. Moreover, this relationship is robust, inter alia, to a number of factors including the introduction of liquidity effects, use of data measured over a different frequency, and market conditions.
Please click on the title to find the paper on the Science Direct website, or click here for a late working paper version.

Economic Valuation of Liquidity Timing. Journal of Banking and Finance (2013) 37:12, p5073-5087
Joint work with Dennis Karstanje, Elvira Sojli and Wing Wah Tham
Abstract: This paper conducts a horse-race of different liquidity proxies using dynamic asset allocation strategies to evaluate the short-horizon predictive ability of liquidity on monthly stock returns. We assess the economic value of the out-of-sample power of empirical models based on different liquidity measures and find three key results: liquidity timing leads to tangible economic gains; a risk-averse investor will pay a high performance fee to switch from a dynamic portfolio strategy based on various liquidity measures to one that conditions on the Zeros measure (Lesmond, Ogden, and Trzcinka, 1999); the Zeros measure outperforms other liquidity measures because of its robustness in extreme market conditions. These findings are stable over time and robust to controlling for existing market return predictors or considering risk-adjusted returns.
Please click on the title to find the paper on the Science Direct website, or click here for a late working paper version.

Forecasting the U.S. Term Structure of Interest Rates Using a Macroeconomic Smooth Dynamic Factor Model. International Journal of Forecasting (2013) 29:4, p676-694
Joint work with Siem Jan Koopman
Abstract: We extend the class of dynamic factor yield curve models for the inclusion of macro-economic factors. We benefit from recent developments in the dynamic factor literature for extracting the common factors from a large panel of macroeconomic series and for estimating the parameters in the model. We include these factors into a dynamic factor model for the yield curve, in which we model the salient structure of the yield curve by imposing smoothness restrictions on the yield factor loadings via cubic spline functions. We carry out a likelihood-based analysis in which we jointly consider a factor model for the yield curve, a factor model for the macroeconomic series, and their dynamic interactions with the latent dynamic factors. We illustrate the methodology by forecasting the U.S. term structure of interest rates. For this empirical study we use a monthly time series panel of unsmoothed Fama-Bliss zero yields for treasuries of different maturities between 1970 and 2009, which we combine with a macro panel of 110 series over the same sample period. We show that the relation between the macroeconomic factors and yield curve data has an intuitive interpretation, and that there is interdependence between the yield and macroeconomic factors. Finally, we perform an extensive out-of-sample forecasting study. Our main conclusion is that macroeconomic variables can lead to more accurate yield curve forecasts.
Please click on the title to find the paper on the Science Direct website, or click here for a late working paper version. A discussion of the paper is also included in the issue.

Customer Order Flow, Intermediaries, and Discovery of the Equilibrium Risk-free Rate. Journal of Financial and Quantitative Analysis (2012) 47:4, p821-849
Joint work with Albert Menkveld and Asani Sarkar
Abstract: Macro announcements change the equilibrium riskfree rate. We find that treasury prices reflect part of the impact instantaneously, but intermediaries rely on their customer order flow after the announcement to discover the full impact. This customer flow informativeness is strongest when analyst macro forecasts are most dispersed. The result holds for 30-year treasury futures trading in both electronic and open-outcry markets. We further show that intermediaries benefit from privately recognizing informed customer flow, as their own-account trading profitability correlates with customer order access.
Please click on the title to find the paper on the Cambridge Journals website, or click here for a late working paper version. Click here for the web appendix.

Maximum Likelihood Estimation for Dynamic Factor Models with Missing Data. Journal of Economic Dynamics and Control (2011) 35:8, p1358-1368
Joint work with Borus Jungbacker and Siem Jan Koopman
Abstract: This paper concerns the maximum likelihood estimation of parameters in a highdimensional dynamic factor model. To accommodate missing data in the analysis, we propose a new model representation for the dynamic factor model. It allows the Kalman filter and related smoothing methods to evaluate the likelihood function and to produce optimal factor estimates in a computationally efficient way when missing data is present. The implementation details of our methods for signal extraction and maximum likelihood estimation are discussed in detail. The computational gains of the new devices are presented based on simulated data-sets with varying numbers of missing entries.
Please click on the title to find the paper on ScienceDirect, or click here for a late working paper version.

Analyzing the Term Structure of Interest Rates using the Dynamic Nelson-Siegel Model with Time-Varying Parameters. Journal of Business and Economic Statistics (2010) 28:3, p329-343
Joint work with Siem Jan Koopman and Max Mallee
Abstract: In this paper we introduce time-varying parameters in the dynamic Nelson-Siegel yield curve model for the simultaneous analysis and forecasting of interest rates of different maturities, known as the term structure. The Nelson-Siegel model has been recently reformulated as a dynamic factor model where the latent factors are interpreted as the level, slope and curvature of the term structure. The factors are modeled jointly as a vector autoregressive process. We propose to extend this framework in two directions. First, the factor loadings in the Nelson-Siegel yield model depend on a single loading parameter. We allow this parameter to be time-varying by treating it as the fourth latent factor that is modeled jointly with the other factors in the vector autoregressive process. Second, we investigate in detail whether the overall volatility in interest rates is constant over time. For this purpose, we introduce a common volatility component that is specified as a GARCH (generalized autoregressive conditional heteroskedasticity) process. The common volatility component is scaled separately for each maturity by an unknown coefficient. We further investigate whether the innovations of the factors are also subject to a common volatility component. Based on a dataset of yield curves that is analyzed by others, we present empirical evidence of considerable increases in within-sample goodness-of-fit when time-varying loadings and volatilities in the dynamic Nelson-Siegel yield model are introduced.
Please click on the title to find the paper on the ASA website, or click here for a late working paper version.