Description
TitleEssays on uncertainty measures and forecasting
Date Created2020
Other Date2020-05 (degree)
Extent1 online resource (xii, 153 pages) : illustrations
DescriptionThe recent decade saw the rapid increase of data size and frequency available for economic and financial analysis. This also brings the opportunity to gain new insights into the interplay between uncertainty, financial markets, and the macro economy, utilizing recent advances in high-frequency financial econometrics, as well as in macroeconometrics.
In Chapter 2, we introduce a class of multi-frequency macroeconomic and financial volatility risk factors. The factors are designed to measure uncertainty, and are latent variables extracted from a state space model that includes multiple different frequencies of non-parametrically estimated components of quadratic variation. When forecasting growth rates of monthly frequency macroeconomic variables, including housing starts, industrial production and nonfarm payroll employment, use of the new risk factors results in significant improvements in predictive performance.
Additionally, when used to forecast corporate yields, the risk factors result in monotonically increasing predictive accuracy gains, as one moves from predicting bonds with higher ratings to predicting bonds with lower ratings. This is consistent with the existence of a natural pricing channel wherein financial risk is more important, predictively, for lower grade bonds. Although the above results are promising, it should be noted that there are exceptions. In particular, we find that when forecasting personal consumption, consumer sentiment and price growth rates, the use of simple daily volatility measures often yield superior predictions. Nevertheless, the preponderance of evidence presented in this paper points to impressive predictive gains associated with the use of the new volatility risk factors. Finally, it is worth noting that a variety of other risk factors, including the Aruoba et al. (2009b) business conditions index as well as a new financial-macroeconomic risk factor based on our multi-frequency approach often also contain marginal predictive content for the variables that we examine, although their inclusion does not reduce the importance of our multi-frequency volatility risk factor.
In Chapter 3, we examine the usefulness of a large variety of machine learning methods for forecasting daily and monthly sector level equity returns. We also examine the usefulness of three new latent risk factors that are designed to capture key forecasting information associated with financial market stress, market uncertainty, and macroeconomic fundamentals. The factors are variously based on the decomposition (using high frequency financial data) of the quadratic covariation between two assets into continuous and jump components, and the extraction of latent factors from mixed frequency state space models populated with nonparametrically estimated components of quadratic variation and/or low frequency macroeconomic data. In addition to constructing predictions using standard machine learning methods such as random forest, gradient boosting, support vector machine learning, penalized regression, and neural networks, among others, we also investigate the predictive performance of a group of hybrid machine learning methods that combine least absolute shrinkage operator and neural network specification methods. Overall, at the monthly frequency, we find that machine learning methods significantly improve forecasting performance, as measured using mean square forecast error (MSFE) and directional predictive accuracy rate (DPAR), relative to the random walk and linear benchmark alternatives. The “best” method is clearly the random forest method, which “wins” in almost all permutations at the monthly frequency, across all of the “target” variables that we predict. It is also worth noting that our hybrid machine learning methods often outperform individual methods, when forecasting daily data, although predictive gains associated with the use of any machine learning method are substantially reduced when forecasting at a daily versus monthly frequency. Finally, the novel uncertainty factors that we build are present in almost all of our “MSFE-best” and directional “accuracy-best” models, suggesting that the risk factors constructed using both high frequency financial data (e.g., 5-minute frequency S&P500 and sector ETF data) and aggregate low frequency macroeconomic data, are useful for predicting returns.
In Chapter 4, we evaluate the development of new tests and methods used in the evaluation of time series forecasts and forecasting models remains as important today as it has been for the last 50 years. Paraphrasing what Sir Clive W.J. Granger (arguably the father of modern day time series forecasting) said in the 1990s at a conference in Svinkloev, Denmark, ‘OK, the model looks like an interesting extension, but can it forecast better than existing models.’ Indeed, the forecast evaluation literature continues to expand, with interesting new tests and methods being developed at a rapid pace. In this chapter, we discuss a selected group of predictive accuracy tests and model selection methods that have been developed in recent years, and that are now widely used in the forecasting literature. We begin by reviewing several tests for comparing the relative forecast accuracy of different models, in the case of point forecasts. We then broaden the scope of our discussion by introducing density-based predictive accuracy tests. We conclude by noting that predictive accuracy is typically assessed in terms of a given loss function, such as mean squared forecast error or mean absolute forecast error. Most tests, including those discussed here, are consequently loss function dependent, and the relative forecast superiority of predictive models is therefore also dependent on specification of a loss function. In light of this fact, we conclude this chapter by discussing loss function robust predictive density accuracy tests that have recently been developed using principles of stochastic dominance.
NotePh.D.
NoteIncludes bibliographical references
Genretheses, ETD doctoral
LanguageEnglish
CollectionSchool of Graduate Studies Electronic Theses and Dissertations
Organization NameRutgers, The State University of New Jersey
RightsThe author owns the copyright to this work.