Prediction and simulation using simple models characterized by nonstationarity and seasonality

Abstract

In this paper, we provide new evidence on the empirical usefulness of various simple seasonal models, and underscore the importance of carefully designing criteria by which one judges alternative models. In particular, we underscore the importance of both choice of forecast or simulation horizon and choice between minimizing point or distribution based loss measures. Our empirical analysis centers around the implementation of a series of simulation and prediction experiments, as well as a discussion of the stochastic properties of seasonal unit root models. Our prediction experiments are based on analysis of a group of 14 variables have been chosen to closely mimic the set of indicators used by the Federal Reserve to help in setting U.S. monetary policy, and our simulation experiments are based on a comparison of simulated and historical distributions of said variables using the testing approach of Corradi and Swanson (2007a). A key impetus for this paper stems from the fact that various financial service companies routinely create "economic scenarios", whereby seasonal and nonstationary financial and economic variables such as those examined here are simulated (and predicted) using relatively simple time series models. These "economic scenarios" are subsequently used in risk management and asset allocation, as is often mandated by various world financial regulatory authorities. Our findings suggest that a simple version of the seasonal unit root (SUROOT) model performs very well for predicting 8 of 14 variables, when the forecast horizon is 1-step ahead. However, for horizons of greater than one-step ahead, our SUROOT model perform poorly when used for prediction, suggesting that parameter estimation error is crucial to understanding the empirical performance of such models. This "parameter estimation error" result is confirmed via a series of Monte Carlo experiments. Simulation experiments yield similar conclusions, although SUROOT models in this case are useful for constructing "forward" conditional distributions at 1-and 3-step ahead horizons. Interestingly, simple periodic autoregressions do not have this property, and are found to perform very well in both prediction and simulation experiments, at all horizons up to 60 months ahead.JEL classification: C13, C22, C52, C53.

Advances in financial risk management and economic policy uncertainty: An overview

“…forecasting value-at-risk using block structure multivariate stochastic volatility models (Asai, Caporin and McAleer, 2014), the time-varying causality between spot and futures crude oil prices: a regime switching approach (Balcilar, Gungor and Hammoudeh, 2014), a regimedependent assessment of the information transmission dynamics between oil prices, precious metal prices and exchange rates (Balcilar, Hammoudeh and Fru Asaba, 2014), a practical approach to constructing price-based funding liquidity factors (Bouwman, Buis, PieterseBloem and Tham, 2014), realized range volatility forecasting: dynamic features and predictive variables (Caporin and Velo, 2014), modelling a latent daily tourism financial conditions index (Chang, 2014), bank ownership, financial segments and the measurement of systemic risk: an application of CoVaR (Drakos and Kouretas, 2014), model-free volatility indexes in the financial literature: a review (Gonzalez-Perez, 2014), robust hedging performance and volatility risk in option markets: application to Standard and Poor's 500 and Taiwan index options (Han, Chang, Kuo and Yu, 2014), price cointegration between sovereign CDS and currency option markets in the financial crises of 2007-2013(Hui and Fong, 2014, whether zombie lending should always be prevented (Jaskowski, 2014), preferences of risk-averse and risk-seeking investors for oil spot and futures before, during and after the global financial crisis (Lean, McAleer and Wong, 2014), managing financial risk in Chinese stock markets: option pricing and modeling under a multivariate threshold autoregression (Li, Ng and Chan, 2014), managing systemic risk in The Netherlands (Liao, Sojli and Tham, 2014), mean-variance portfolio methods for energy policy risk management (Marrero, Puch and Ramos-Real, 2014), on robust properties of the SIML estimation of volatility under micro-market noise and random sampling (Misaki and Kunitomo, 2014), ALRIGHT: Asymmetric LaRge-Scale (I)GARCH with Hetero-Tails (Paolella and Polak, 2014), the economic fundamentals and economic policy uncertainty of Mainland China and their impacts on Taiwan and Hong Kong (Sin, 2014), prediction and simulation using simple models characterized by nonstationarity and seasonality 5 (Swanson and Urbach, 2014), and volatility forecast of stock indexes by model averaging using high frequency data (Wang and Nishiyama, 2014).…”

Visual Data Analysis and Simulation Prediction for COVID-19 in Saudi Arabia Using SEIR Prediction Model

“…Simulation is another substitute in accepting business complications, forecasting upcoming trends, and endorsing best decisions. They explained the basics of simulation models and point out three of its advantages [19,20]. Table 1 shows the dataset applied for the outcomes.…”

Forecasting and nowcasting emerging market GDP growth rates: The role of latent global economic policy uncertainty and macroeconomic data surprise factors

“…5 . For further discussion of the importance of seasonality and issues associated with seasonal adjustment, including the fact that use of YoY growth rates may induce spurious correlation, see Ghysels, Osborn, and Sargent (2001), sues associated with seasonal adjustment, including the fact that use of YoY growth rates may induce spurious correlation, see Ghysels et al (2001), Luciani et al (2018), Swanson and Urbach (2015), and Uhlig (2009).…”