Castellani M.,University of Bergen |
Dos Santos E.A.,Banco de Portugal
Studies in Computational Intelligence | Year: 2014
This chapter investigates the use of different artificial intelligence and classical techniques for forecasting the monthly yield of the US 10-year Treasury bonds from a set of four economic indicators. The task is particularly challenging due to the sparseness of the data samples and the complex interactions amongst the variables. At the same time, it is of high significance because of the important and paradigmatic role played by the US market in the world economy. Four data-driven artificial intelligence approaches are considered: a manually built fuzzy logic model, a machine learned fuzzy logic model, a self-organising map model, and a multi-layer perceptron model. Their prediction accuracy is compared with that of two classical approaches: a statistical ARIMA model and an econometric error correction model. The algorithms are evaluated on a complete series of end-month US 10-year Treasury bonds yields and economic indicators from 1986:1 to 2004:12. In terms of prediction accuracy and reliability, the best results are obtained by the three parametric regression algorithms, namely the econometric, the statistical, and the multi-layer perceptron model. Due to the sparseness of the learning data samples, the manual and the automatic fuzzy logic approaches fail to follow with adequate precision the range of variations of the US 10-year Treasury bonds. For similar reasons, the self-organising map model performs unsatisfactorily. Analysis of the results indicates that the econometric model has a slight edge over the statistical and the multi-layer perceptron models. This suggests that pure data-driven induction may not fully capture the complicated mechanisms ruling the changes in interest rates. Overall, the prediction accuracy of the best models is only marginally better than the prediction accuracy of a basic one-step lag predictor. This result highlights the difficulty of the modelling task and, in general, the difficulty of building reliable predictors for financial markets. © 2014 Springer International Publishing Switzerland.
Moreira S.,Banco de Portugal |
Barros P.P.,New University of Lisbon
Health Economics | Year: 2010
Double health insurance coverage exists when an individual benefits from more than one health insurance plan at the same time. We examine the impact of such supplementary insurance on the utilisation of doctor consultations in Portugal, taking advantage of institutional features which make double coverage plausibly exogenous. The novelty is that the analysis is carried out for different points of the conditional distribution, not only for its mean location, within the context of count data modelling and without imposing restrictive parametric assumptions. Results indicate that double coverage creates additional utilisation of health care across the whole outcome distribution for both public and private second layers of health insurance coverage but with greater magnitude in the latter group. We unveil that this additional consumption effect is relatively smaller for more frequent users. Copyright © 2010 John Wiley & Sons, Ltd.
Rodrigues P.M.M.,Banco de Portugal |
Salish N.,University of Bonn
Journal of Cardiovascular Translational Research | Year: 2015
This paper proposes threshold models to analyze and forecast interval-valued time series. A relatively simple algorithm is proposed to obtain least square estimates of the threshold and slope parameters. The construction of forecasts based on the proposed model and methods for the analysis of their forecast performance are also introduced and discussed, as well as forecasting procedures based on the combination of different models. To illustrate the usefulness of the proposed methods, an empirical application on a weekly sample of S&P500 index returns is provided. The results obtained are encouraging and compare very favorably to available procedures. © 2014, Springer-Verlag Berlin Heidelberg.
Dias F.,Banco de Portugal |
Pinheiro M.,Banco de Portugal |
Pinheiro M.,University of Lisbon |
Rua A.,Banco de Portugal
Journal of Forecasting | Year: 2010
The simplicity of the standard diffusion index model of Stock and Watson has certainly contributed to its success among practitioners, resulting in a growing body of literature on factor-augmented forecasts. However, as pointed out by Bai and Ng, the ranked factors considered in the forecasting equation depend neither on the variable to be forecast nor on the forecasting horizon. We propose a refinement of the standard approach that retains the computational simplicity while coping with this limitation. Our approach consists of generating a weighted average of all the principal components, the weights depending both on the eigenvalues of the sample correlation matrix and on the covariance between the estimated factor and the targeted variable at the relevant horizon. This 'targeted diffusion index' approach is applied to US data and the results show that it outperforms considerably the standard approach in forecasting several major macroeconomic series. Moreover, the improvement is more significant in the final part of the forecasting evaluation period. © 2009 John Wiley and Sons, Ltd.
Perelman J.,New University of Lisbon |
Felix S.,Banco de Portugal |
Felix S.,New University of Lisbon |
Santana R.,New University of Lisbon
Health Policy | Year: 2015
The Great Recession started in Portugal in 2009, coupled with severe austerity. This study examines its impact on hospital care utilization, interpreted as caused by demand-side effects (related to variations in population income and health) and supply-side effects (related to hospitals' tighter budgets and reduced capacity).The database included all in-patient stays at all Portuguese NHS hospitals over the 2001-2012 period (n= 17.7 millions). We analyzed changes in discharge rates, casemix index, and length of stay (LOS), using a before-after methodology. We additionally measured the association of health care indicators to unemployment.A 3.2% higher rate of discharges was observed after 2009. Urgent stays increased by 2.5%, while elective in-patient stays decreased by 1.4% after 2011. The LOS was 2.8% shorter after the crisis onset, essentially driven by the 4.5% decrease among non-elective stays. A one percentage point increase in unemployment rate was associated to a 0.4% increase in total volume, a 2.3% decrease in day cases, and a 0.1% decrease in LOS.The increase in total and urgent cases may reflect delayed out-patient care and health deterioration; the reduced volume of elective stays possibly signal a reduced capacity; finally, the shorter stays may indicate either efficiency-enhancing measures or reduced quality. © 2015 Elsevier Ireland Ltd.