|
on Forecasting |
| By: | Riccardo LUCCHETTI (Universita' Politecnica delle Marche, Dipartimento di Economia); Giulio PALOMBA ([n.d.]) |
| Abstract: | Forecasting models for bond yields often use macro data to improve their properties. Unfortunately, macro data are not available at frequencies higher than monthly.;In order to mitigate this problem, we propose a nonlinear VEC model with conditional heteroskedasticity (NECH) and find that such model has superior in-sample performance than models which fail to encompass nonlinearities and/or GARCH-type effects.;Out-of-sample forecasts by our model are marginally superior to competing models; however, the data points we used for evaluating forecasts refer to a period of relative tranquillity on the financial markets, whereas we argue that our model should display superior performance under "unusual" circumstances. |
| Keywords: | conditional heteroskedasticity, forecasting, interest rates, nonlinear cointegration |
| JEL: | C32 C53 E43 |
| Date: | 2006–05 |
| URL: | http://d.repec.org/n?u=RePEc:anc:wpaper:261&r=for |
| By: | Ilias Lekkos (Eurobank Ergasias); Costas Milas (Keele University, Department of Economics); Theodore Panagiotidis (Loughborough University) |
| Abstract: | This paper explores the ability of common risk factors to predict the dynamics of US and UK interest rate swap spreads within a linear and a non-linear framework. We reject linearity for the US and UK swap spreads in favour of a regime-switching smooth transition vector autoregressive (STVAR) model, where the switching between regimes is controlled by the slope of the US term structure of interest rates. The first regime is characterised by a "flat" term structure of US interest rates, while the alternative is characterised by an "upward" sloping US term structure. We compare the ability of the STVAR model to predict swap spreads with that of a non-linear nearest-neibours model as well as that of linear AR and VAR models. We find some evidence that the nearest-neighbours and STVAR models predict better than the linear AR and VAR models. However, the evidence is not overwhelming as it is sensitive to swap spread maturity. We also find that within the non-linear class of models, the nearest-neighbours model predicts better than the STVAR model US swap spreads in periods of increasing risk conditions and UK swap spreads in periods of decreasing risk conditions. |
| Keywords: | Interest rate swap spreads, term structure of interest rates, regime switching, smooth transition models, nearest-neighbours, forecasting |
| JEL: | C51 C52 C53 E43 |
| Date: | 2005–02 |
| URL: | http://d.repec.org/n?u=RePEc:kee:kerpuk:2005/13&r=for |
| By: | Kevin J. Lansing |
| Abstract: | This paper introduces a form of boundedly-rational expectations into an otherwise standard New-Keynesian Phillips curve. The representative agent's forecast rule is optimal (in the sense of minimizing mean squared forecast errors), conditional on a perceived law of motion for inflation and observed moments of the inflation time series. The perceived law of motion allows for both temporary and permanent shocks to inflation, the latter intended to capture the possibility of evolving shifts in the central bank's inflation target. In this case, the agent's optimal forecast rule defined by the Kalman filter coincides with adaptive expectations, as shown originally by Muth (1960). I show that the perceived optimal value of the gain parameter assigned to the last observed inflation rate is given by the fixed point of a nonlinear map that relates the gain parameter to the autocorrelation of inflation changes. The model allows for either a constant gain or variable gain, depending on the length of the sample period used by the agent to compute the autocorrelation of inflation changes. In the variable-gain setup, the equilibrium law of motion for inflation is nonlinear and can generate time-varying inflation dynamics similar to those observed in long-run U.S. data. The model's inflation dynamics are driven solely by white-noise fundamental shocks propagated via the expectations feedback mechanism; all monetary policy-dependent parameters are held constant. |
| Keywords: | Inflation (Finance) ; Phillips curve ; Econometric models |
| Date: | 2006 |
| URL: | http://d.repec.org/n?u=RePEc:fip:fedfwp:2006-15&r=for |
| By: | Andrew Ang; Geert Bekaert; Min Wei |
| Abstract: | Surveys do! We examine the forecasting power of four alternative methods of forecasting U.S. inflation out-of-sample: time series ARIMA models; regressions using real activity measures motivated from the Phillips curve; term structure models that include linear, non-linear, and arbitrage-free specifications; and survey-based measures. We also investigate several methods of combining forecasts. Our results show that surveys outperform the other forecasting methods and that the term structure specifications perform relatively poorly. We find little evidence that combining forecasts produces superior forecasts to survey information alone. When combining forecasts, the data consistently places the highest weights on survey information. |
| Date: | 2006 |
| URL: | http://d.repec.org/n?u=RePEc:fip:fedgfe:2006-15&r=for |
| By: | Juan Ignacio Pena; Rosa Rodriguez |
| Abstract: | This paper presents a model linking two financial markets (stocks and bonds) with the real business cycle, in the framework of the Consumption Capital Asset Pricing Model with Generalized Isoelastic Preferences. Besides interest rate term spread, the model includes a new variable to forecast economic activity: stock market term spread, which constitutes the slope of expected stock market returns. The empirical evidence documented in this paper suggests systematic relationships between the state of the business cycle and the shapes of two yield curves (interest rates and expected stock returns). Results are robust to changes in measures of economic growth, stock prices, interest rates and expectation-generating mechanisms. |
| Date: | 2006–05 |
| URL: | http://d.repec.org/n?u=RePEc:cte:wbrepe:wb063209&r=for |
| By: | NA (NA) |
| Abstract: | This paper presents a study of errors in forecasting the population of Metropolitan Statistical Areas and the Primary MSAs of Consolidated Metropolitan Statistical Areas and New England MAs. The forecasts are for the year 2000 and are based on a semi-structural model estimated by Mills and Lubelle using 1970 to 1990 census data on population, employment and relative real wages. This model allows the testing of regional effects on population and employment growth. The year 2000 forecasts are for 321 MSAs as they were defined in 1990. Actual year 2000 populations for these MSAs are constructed using the MSA components lists for 1990. Forecast errors are constructed for these “historic” MSAs. The forecast errors for the entire set of cities are examined for regional patterns. A subset of 77 cities is examined more carefully using the State of the Nations Cities (SONC) data base prepared by the Center for Urban Policy Research. SONC contains observations on 2000 demographic and socioeconomic variables for all 77 MSAs in the data set. Selected variables will be used to test a model of forecast areas developed for this project to determine if there are systematic relationships between selected variables and the forecast errors and to determine if a semi-structural model based on urban characteristics variables can improve urban population forecasts. |
| Date: | 2005–12 |
| URL: | http://d.repec.org/n?u=RePEc:bsu:wpaper:200510&r=for |
| By: | Kirill F. Andreev (Max Planck Institute for Demographic Research, Rostock, Germany); James W. Vaupel (Max Planck Institute for Demographic Research, Rostock, Germany) |
| Abstract: | - |
| JEL: | J1 Z0 |
| Date: | 2006–05 |
| URL: | http://d.repec.org/n?u=RePEc:dem:wpaper:wp-2006-012&r=for |
| By: | Oliver Lipps; Frank Betz (Mannheim Research Institute for the Economics of Aging (MEA)) |
| Abstract: | This contribution builds upon a former paper by the authors (Lipps and Betz 2004), in which a stochastic population projection for East- and West Germany is performed. Aim was to forecast relevant population parameters and their distribution in a consistent way. We now present some modifications, which have been modelled since. First, population parameters for the entire German population are modelled. In order to overcome the modelling problem of the structural break in the East during reunification, we show that the adaptation process of the relevant figures by the East can be considered to be completed by now. As a consequence, German parameters can be modelled just by using the West German historic patterns, with the start-off population of entire Germany. Second, a new model to simulate age specific fertility rates is presented, based on a quadratic spline approach. This offers a higher flexibility to model various age specific fertility curves. The simulation results are compared with the scenario based official forecasts for Germany in 2050. Exemplary for some population parameters (e.g. dependency ratio), it can be shown that the range spanned by the medium and extreme variants correspond to the s -intervals in the stochastic framework. It seems therefore more appropriate to treat this range as a s-interval covering about two thirds of the true distribution. |
| JEL: | Z00 |
| Date: | 2004–09–22 |
| URL: | http://d.repec.org/n?u=RePEc:xrs:meawpa:04059&r=for |