In this issue we haven Exchange Rate Equations Based on Interest Rate Rules, Forecasting Wheat Output and Profits from Cropping Systems Using Simulation Models, Evaluation of Wavelet-based Core Inflation Measures, Measuring overfitting and mispecification in nonlinear models, Prediction Markets to Forecast Electricity Demand, and more.
|
Date:
|
2011
|
|
By:
|
Mahir Binici
Yin-Wong Cheung
|
|
URL:
|
|
|
Using exchange rate data on five currencies vis-à-vis the US dollar, this paper examines the insample and out-of-sample performance of exchange rate equations derived from alternative empirical and optimal interest rate rules. These rules could have either homogeneous or heterogeneous response coefficients. Our exercise shows that these exchange rate equations do not offer good in-sample explanatory power consistently across currencies and over time. The relative forecasting performance of these exchange rate equations tend to vary across currencies and over time and bears limited relationship with the relative in-sample performance. When the forecast performance is compared with a random walk model, these exchange rate equations offer no better performance under the usual MSFE criterion but are better when the ability of predicting the direction of change is considered.
|
|
|
Keywords:
|
Taylor Rule, Exchange Rate Determination, Forecast Comparison, Mean Squared Forecast Error, Direction of Change
|
|
JEL:
|
F31
|
|
Date:
|
2011-07
|
|
By:
|
Bhattacharya, Rudrani (National Institute of Public Finance and Policy)
Pandey, Radhika (National Institute of Public Finance and Policy) Veronese, Giovanni (Bank of Italy)
|
|
URL:
|
|
|
Tracking growth in the Indian economy would be best performed using a measure like GDP. Unfortunately official estimates of this indicator are released with quarterly frequency and with considerable delay. This paper compares different approaches to the short term forecasting (nowcasting) of real GDP growth in India and evaluates methods to optimally gauge the current state of the economy. Univariate quarterly models are compared with bridge models that exploit the available monthly indicators containing information on current quarter developments. In the forecasting exercise we perform a pseudo real-time simulation: by properly taking into account the actual publication lags of the series, we replicate the information set available to the policymaker at each point of time. We find that bridge models perform satisfactorily in predicting current quarter GDP growth. This result follows from the actual estimation technique used to construct the official quarterly national accounts, still largely dependent on a narrow information set. Our analysis also suggests mixed evidences about the additional predictive power of Indian survey data with respect to the hard data already used in the national accounts.
|
|
|
Keywords:
|
Nowcasting ; Bridge model ; Factor model ; Emerging markets ; India
|
|
JEL:
|
C22
|
|
Date:
|
2011-07-25
|
|
By:
|
Kwasnicki, Witold
|
|
URL:
|
|
|
In the first part of the paper an overview of the long-term global economic growth forecasts is presented (e.g., forecasts of Uri Dadush and Bennett Stancil of the Carnegie Endowment for International Peace, a report by HSBC, CitiGroup report, reports of PricewaterhouseCoopers, or Goldman Sachs reports). In this context, the diversified views and opinions on future economic development of China and India (currently considered as the fastest-growing major economies in the world) are presented. In the second part of the article, an extrapolative forecast of global GDP and an estimation of the economies of China and India in global production by 2050 based on the so-called evolutionary model of competition is outlined. The evolutionary model of competition enables to estimate the competitiveness of national economies, therefore in the second part of the paper we presents also the results of estimation of the competitiveness of the economies of India and China after World War II. One aim of that research is to compare the competitiveness of China and India with the leaders of economic development in the twentieth century, namely the United States, Great Britain, Germany, Japan and the European Union. The summary of these considerations are estimations the shares of GDP of China and India in the global product based on global scenarios of the competitiveness changes of these economies over the next 40 years.
|
|
|
Keywords:
|
future studies; forecasting; globalization; economic growth
|
|
JEL:
|
O11
|
|
Date:
|
2010-09
|
|
By:
|
Odwori, P.O.
Mapelu, M.Z. Odhiambo, Mark O. Nyangweso, P.M.
|
|
URL:
|
|
|
Simulation models have been used successfully to forecast productivity of cropping systems under various weather, management and policy scenarios. These models have helped farmers make efficient resource allocation decisions. However, in Kenya simulation models have not been used extensively and more specifically in modeling maize cropping system. The study aimed at forecasting productivity and profitability of maize cropping system in Uasin Gishu district, Kenya. Both primary and secondary data were used. Both time series and cross-sectional data for variables of interest were collected and complemented by a survey of 20 maize farmers who were systematically selected to verify information obtained from secondary sources. Cropping Systems simulation model and Monte Carlo simulation were used to determine maize output and profits under alternative price scenarios. Even though, simulated yields underestimated actual maize yield both at the district and across the four agro-ecological zones, the deviation from the actual yield was marginal. It is recommended that Cropsyst and Monte Carlo models be included among a bundle of tools for decision making. Further research is also required to test the two models under different locations, soil types, management styles and scales of production.
|
|
|
Keywords:
|
Forecasting, Yields, Profits, Maize cropping system, Simulation models, Crop Production/Industries,
|
|
Date:
|
2010-09
|
|
By:
|
Nyangweso, P.M.
Odmori, Paul Okelo Mapelu, M.Z. Odhiambo, Mark O.
|
|
URL:
|
|
|
Simulation models have been used successfully to forecast productivity of cropping systems under various weather, management and policy scenarios. These models have helped farmers make efficient resource allocation decisions. However, in Kenya simulation models have not been used extensively and more specifically in modeling large scale cropping systems. The study aimed at forecasting productivity and profitability of wheat cropping systems in Uasin Gishu district, Kenya. Both primary and secondary data were used. Both time series and cross-sectional data for variables of interest were collected and complemented by a survey of 20 wheat farmers who were systematically selected to verify information obtained from secondary sources. Cropping Systems simulation model and Monte Carlo simulation were used to determine wheat output and profits under alternative price scenarios. Even though, simulated yields overestimated actual field wheat yield both at the district and across the four agro-ecological zones, the deviation from the actual field yield was marginal. It is recommended that Cropsyst and Monte Carlo models be included among a bundle of tools for decision making. Further research is also required to test the two models under different locations, diverse soil types, varied management styles and different scales of production.
|
|
|
Keywords:
|
Wheat, cropping system, simulation, forecasting, productivity, profits, Crop Production/Industries,
|
|
Date:
|
2011
|
|
By:
|
Erick Lahura (Departamento de Economía – Pontificia Universidad Católica del Perú)
Marco Vega (Departamento de Economía – Pontificia Universidad Católica del Perú)
|
|
URL:
|
|
|
Under inflation targeting and other related monetary policy regimes, the identication of non-transitory inflation and forecasts about future inflation constitute key ingredients for monetary policy decisions. In practice, central banks perform these tasks using so-called “core inflation measures”. In this paper we construct alternative core inflation measures using wavelet functions and multiresolution analysis (MRA), and then evaluate their relevance for monetary policy. The construction of wavelet-based core inflation measures (WIMs) is relatively new in the literature and their assessment has not been addressed formally, this paper being the first attempt to perform both tasks for the case of Peru. Another main contribution of this paper is that it proposes two alternative criteria for evaluating core inflation measures: (i) a VAR-based long-run criterion, and (ii) forecast-based criteria. Evidence from Peru shows that WIMs are superior in terms of long-run performance, and that they could improve short-term (up-to-6-months) inflation forecasts.
|
|
|
Keywords:
|
Core inflation, wavelets, forecast, structural VAR
|
|
JEL:
|
C45
|
|
Date:
|
2011-08
|
|
By:
|
Bilger M.;
Manning W.G;
|
|
URL:
|
|
|
We start by proposing a new measure of overfitting expressed on the untransformed scale of the dependent variable, which is generally the scale of interest to the analyst.We then show that with nonlinear models shrinkage due to overfitting gets confounded by shrinkage—or expansion— arising from model misspecification. Out-of-sample predictive calibration can in fact be expressed as in-sample calibration times 1 minus this new measure of overfitting. We finally argue that re-calibration should be performed on the scale of interest and provide both a simulation study and a real-data illustration based on health care expenditure data.
|
|
|
Keywords:
|
overfitting, shrinkage, misspecification, forecasting, health care expenditure
|
|
Date:
|
2010-10-01
|
|
By:
|
Nabil I. Al-Najjar
Luciano De Castro
|
|
URL:
|
|
|
A preference is invariant with respect to a transformation tau if its ranking of acts is unaffected by a reshuffling of the states under tau. We show that any invariant preference must be parametric: there is a unique sufficient set of parameters such that the preference ranks acts according to their expected utility given the parameters. This property holds for all non-trivial preferences, provided only that they are reflexive, transitive, monotone, continuous and mixture linear.
|
|
Taken from the NEP-FOR mailing list edited by Rob Hyndman.