|
on Forecasting |
| By: | David Jamieson Bolder; Yuliya Romanyuk |
| Abstract: | Model risk is a constant danger for financial economists using interest-rate forecasts for the purposes of monetary policy analysis, portfolio allocations, or risk-management decisions. Use of multiple models does not necessarily solve the problem as it greatly increases the work required and still leaves the question "which model forecast should one use?" Simply put, structural shifts or regime changes (not to mention possible model misspecifications) make it difficult for any single model to capture all trends in the data and to dominate all alternative approaches. To address this issue, we examine various techniques for combining or averaging alternative models in the context of forecasting the Canadian term structure of interest rates using both yield and macroeconomic data. Following Bolder and Liu (2007), we study alternative implementations of four empirical term structure models: this includes the Diebold and Li (2003) approach and three associated generalizations. The analysis is performed using more than 400 months of data ranging from January 1973 to July 2007. We examine a number of model-averaging schemes in both frequentist and Bayesian settings, both following the literature in this field (such as de Pooter, Ravazzolo and van Dijk (2007)) in addition to introducing some new combination approaches. The forecasts from individual models and combination schemes are evaluated in a number of ways; preliminary results show that model averaging generally assists in mitigating model risk, and that simple combination schemes tend to outperform their more complex counterparts. Such findings carry significant implications for central-banking analysis: a unified approach towards accounting for model uncertainty can lead to improved forecasts and, consequently, better decisions. |
| Keywords: | Interest rates; Econometric and statistical methods |
| JEL: | C11 E43 E47 |
| Date: | 2008 |
| URL: | http://d.repec.org/n?u=RePEc:bca:bocawp:08-34&r=for |
| By: | Andrea Cipollini; Giuseppe Missaglia |
| Abstract: | In this paper, using industry sector stock returns as proxies of firm asset values, we obtain bank capital requirements (through the cycle). This is achieved by Montecarlo simulation of a bank loan portfolio loss density. We depart from the Basel 2 analytical formula developed by Gordy (2003) for the computation of the economic capital by, first, allowing dynamic heterogeneity in the factor loadings, and, also, by accounting for stochastic dependent recoveries. Dynamic heterogeneity in the factor loadings is introduced by using dynamic forecast of a Dynamic Factor model fitted to a large dataset of macroeconomic credit drivers. The empirical findings show that there is a decrease in the degree of Portfolio Credit Risk, once we move from the Basel 2 analytic formula to the Dynamic Factor model specification. |
| Keywords: | Dynamic Factor Model, Forecasting, Stochastic Simulation, Risk Management, Banking |
| JEL: | C32 C53 E17 G21 G33 |
| Date: | 2008–02 |
| URL: | http://d.repec.org/n?u=RePEc:mod:recent:010&r=for |
| By: | Andrea Cipollini; George Kapetanios |
| Abstract: | In this paper we use principal components analysis to obtain vulnerability indicators able to predict financial turmoil. Probit modelling through principal components and also stochastic simulation of a Dynamic Factor model are used to produce the corresponding probability forecasts regarding the currency crisis events a®ecting a number of East Asian countries during the 1997-1998 period. The principal components model improves upon a number of competing models, in terms of out-of-sample forecasting performance. |
| Keywords: | Financial Contagion, Dynamic Factor Model |
| JEL: | C32 C51 F34 |
| Date: | 2008–03 |
| URL: | http://d.repec.org/n?u=RePEc:mod:recent:014&r=for |
| By: | Jonas Dovern; Johannes Weisser |
| Abstract: | In this paper, we use survey data to analyze the rationality of professional macroeconomic forecasts. We analyze both individual forecasts and average forecasts. We provide evidence on the properties of forecasts for all the G7-counties and four different macroeconomic variables. Furthermore, we present a modification to the structural model which is commonly used to model the forecast errors of fixed event forecasts in the literature. Our results confirm that average forecasts should be used with caution, since even if all individual forecasts are rational the hypothesis of rationality is often rejected by the aggregate forecasts. We find that there are not only large differences in the performance of forecasters across countries but also across different macroeconomic variables; in general, forecasts tend to be biased in situations where forecasters have to learn about large structural shocks or gradual changes in the trend of a variable |
| Keywords: | Evaluating forecasts,Macroeconomic Forecasting,Rationality,Survey Data,Fixed-Event Forecasts |
| JEL: | C25 E32 E37 |
| Date: | 2008–09 |
| URL: | http://d.repec.org/n?u=RePEc:kie:kieliw:1447&r=for |
| By: | Andrew Hodge (Reserve Bank of Australia); Tim Robinson (Reserve Bank of Australia); Robyn Stuart (Reserve Bank of Australia) |
| Abstract: | This paper estimates a small structural model of the Australian economy, designed principally for forecasting the key macroeconomic variables of output growth, underlying inflation and the cash rate. In contrast to models with purely statistical foundations, which are often used for forecasting, the Bayesian Vector Autoregressive Dynamic Stochastic General Equilibrium (BVAR-DSGE) model uses the theoretical information of a DSGE model to offset in-sample over-fitting. We follow the method of Del Negro and Schorfheide (2004) and use a variant of the small open economy DSGE model of Lubik and Schorfheide (2007) to provide prior information for the VAR. The forecasting performance of the model is competitive with benchmark models such as a Minnesota VAR and an independently estimated DSGE model. |
| Keywords: | BVAR-DSGE; forecasting |
| JEL: | C11 C53 E37 |
| Date: | 2008–09 |
| URL: | http://d.repec.org/n?u=RePEc:rba:rbardp:rdp2008-04&r=for |
| By: | Frank Schorfheide; Keith Sill; Maxym Kryshko |
| Abstract: | This paper develops and illustrates a simple method to generate a DSGE model-based forecast for variables that do not explicitly appear in the model (non-core variables). The authors use auxiliary regressions that resemble measurement equations in a dynamic factor model to link the non-core variables to the state variables of the DSGE model. Predictions for the non-core variables are obtained by applying their measurement equations to DSGE model- generated forecasts of the state variables. Using a medium-scale New Keynesian DSGE model, the authors apply their approach to generate and evaluate recursive forecasts for PCE inflation, core PCE inflation, and the unemployment rate along with predictions for the seven variables that have been used to estimate the DSGE model. |
| Date: | 2008 |
| URL: | http://d.repec.org/n?u=RePEc:fip:fedpwp:08-17&r=for |
| By: | Proietti, Tommaso |
| Abstract: | The paper focuses on the comparison of the direct and iterated AR predictors when Xt is a difference stationary process. In particular, it provides some useful results for comparing the efficiency of the two predictors and for extracting the trend from macroeconomic time series using the two methods. The main results are based on an encompassing representation for the two predictors which enables to derive their properties quite easily under a maintained model. The paper provides an analytic expression for the mean square forecast error of the two predictors and derives useful recursive formulae for computing the direct and iterated coefficients. From the empirical standpoint, we propose estimators of the AR coefficients based on the tapered Yule-Walker estimates; we also provide a test of equal forecast accuracy which is very simple to implement and whose critical values can be obtained with the bootstrap method. Since multistep prediction is tightly bound up with the estimation of the long run component in a time series, we turn to the role of the direct method for trend estimation and derive the corresponding multistep Beveridge-Nelson decomposition. |
| Keywords: | Beveridge-Nelson decomposition; Multistep estimation; Tapered Yule-Walker estimates; Forecast combination. |
| JEL: | C51 E32 C53 E31 C22 |
| Date: | 2008–10–01 |
| URL: | http://d.repec.org/n?u=RePEc:pra:mprapa:10859&r=for |