|
on Forecasting |
| By: | Laura Coroneo; Fabrizio Iacone; Alessia Paccagnini; Paulo Santos Monteiro |
| Abstract: | We test the predictive accuracy of forecasts for the number of COVID-19 fatalities produced by several forecasting teams and collected by the United States Centers for Disease Control and Prevention (CDC), both at the national and state levels. We find three main results. First, at short-horizon (1-week ahead) no forecasting team outperforms a simple time-series benchmark. Second, at longer horizons (3 and 4-weeks ahead) forecasters are more successful and sometimes outperform the benchmark. Third, one of the best performing forecasts is the Ensemble forecast, that combines all available forecasts using uniform weights. In view of these results, collecting a wide range of forecasts and combining them in an ensemble forecast may be a safer approach for health authorities, rather than relying on a small number of forecasts. |
| Keywords: | Forecast evaluation, Forecasting tests, Epidemic. |
| JEL: | C12 C53 I18 |
| Date: | 2020–10 |
| URL: | http://d.repec.org/n?u=RePEc:yor:yorken:20/10&r=all |
| By: | Nadja Klein; Michael Stanley Smith; David J. Nott |
| Abstract: | Recurrent neural networks (RNNs) with rich feature vectors of past values can provide accurate point forecasts for series that exhibit complex serial dependence. We propose two approaches to constructing deep time series probabilistic models based on a variant of RNN called an echo state network (ESN). The first is where the output layer of the ESN has stochastic disturbances and a shrinkage prior for additional regularization. The second approach employs the implicit copula of an ESN with Gaussian disturbances, which is a deep copula process on the feature space. Combining this copula with a non-parametrically estimated marginal distribution produces a deep distributional time series model. The resulting probabilistic forecasts are deep functions of the feature vector and also marginally calibrated. In both approaches, Bayesian Markov chain Monte Carlo methods are used to estimate the models and compute forecasts. The proposed deep time series models are suitable for the complex task of forecasting intraday electricity prices. Using data from the Australian National Electricity Market, we show that our models provide accurate probabilistic price forecasts. Moreover, the models provide a flexible framework for incorporating probabilistic forecasts of electricity demand as additional features. We demonstrate that doing so in the deep distributional time series model in particular, increases price forecast accuracy substantially. |
| Date: | 2020–10 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2010.01844&r=all |
| By: | Valerio Ercolani (Bank of Italy); Filippo Natoli (Bank of Italy) |
| Abstract: | This paper highlights the role of macroeconomic and financial uncertainty in predicting US recessions. In-sample forecasts using probit models indicate that these two variables are the best predictors of recessions at short horizons. Macroeconomic uncertainty has the highest predictive power up to 7 months ahead and becomes the second best predictor --- after the yield curve slope --- at longer horizons. Using data up to end-2018, out-of-sample forecasts show that uncertainty contributed significantly to lowering the probability of a recession in 2019, which indeed did not occur. |
| Keywords: | macroeconomic and financial uncertainty, yield curve slope, recession, probit forecasting model. |
| JEL: | D81 E32 E37 E44 |
| Date: | 2020–09 |
| URL: | http://d.repec.org/n?u=RePEc:bdi:wptemi:td_1299_20&r=all |
| By: | Mukherjee, Paramita; Coondoo, Dipankor; Lahiri, Poulomi |
| Abstract: | In this paper, an attempt has been made to forecast the hourly electricity spot prices in India as this is very important for the bidders in the energy exchange for participating in the day-ahead market. Forecasting high frequency data is a challenging task. In forecasting, different variants of ARMA, ARMA-GARCH models are applied in different contexts, but no unequivocal dominance of a particular model exists. In this paper, based on hourly data for several years for all the regions in India, several variants of ARMAX models are estimated, by combining static and dynamic forecasts. Along with ARMA, intra-day, inter-day and hourly variations in prices as well as seasonalities on weekdays, holidays and festive days are incorporated. ARMAX models in this context performed quite well for forecasting horizons of hourly prices of upto 5 days. Interestingly, the ARMAX models provide reasonably good forecasts for day-ahead-market and the simple structure can be quite easily implemented. Such forecasts are not only essential for the players in the spot market, but also provides insights for policymakers as it reveals several aspects of Indian electricity market including the different dimensions of seasonality in demand. |
| Keywords: | Forecasting, electricity, hourly data, energy, spot price, ARMAX model, day-ahead market |
| JEL: | C53 Q47 |
| Date: | 2019 |
| URL: | http://d.repec.org/n?u=RePEc:pra:mprapa:103161&r=all |
| By: | Kasun Chandrarathna; Arman Edalati; AhmadReza Fourozan tabar |
| Abstract: | By significant improvements in modern electrical systems, planning for unit commitment and power dispatching of them are two big concerns between the researchers. Short-term load forecasting plays a significant role in planning and dispatching them. In recent years, numerous works have been done on Short-term load forecasting. Having an accurate model for predicting the load can be beneficial for optimizing the electrical sources and protecting energy. Several models such as Artificial Intelligence and Statistics model have been used to improve the accuracy of load forecasting. Among the statistics models, time series models show a great performance. In this paper, an Autoregressive integrated moving average (SARIMA) - generalized autoregressive conditional heteroskedasticity (GARCH) model as a powerful tool for modeling the conditional mean and volatility of time series with the T-student Distribution is used to forecast electric load in short period of time. The attained model is compared with the ARIMA model with Normal Distribution. Finally, the effectiveness of the proposed approach is validated by applying real electric load data from the Electric Reliability Council of Texas (ERCOT). KEYWORDS: Electricity load, Forecasting, Econometrics Time Series Forecasting, SARIMA |
| Date: | 2020–09 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2009.13595&r=all |
| By: | Foltas, Alexander; Pierdzioch, Christian |
| Abstract: | We use quantile random forests (QRF) to study the efficiency of the growth forecasts published by three leading German economic research institutes for the sample period from 1970 to 2017. To this end, we use a large array of predictors, including topics extracted by means of computational-linguistics tools from the business-cycle reports of the institutes, to model the information set of the institutes. We use this array of predictors to estimate the quantiles of the conditional distribution of the forecast errors made by the institutes, and then fit a skewed t-distribution to the estimated quantiles. We use the resulting density forecasts to compute the log probability score of the predicted forecast errors. Based on an extensive insample and out-of-sample analysis, we find evidence, particularly in the case of longer-term forecasts, against the null hypothesis of strongly efficient forecasts. We cannot reject weak efficiency of forecasts. |
| Keywords: | Growth forecasts,Forecast efficiency,Quantile-random forests,Density forecasts |
| JEL: | C53 E32 E37 |
| Date: | 2020 |
| URL: | http://d.repec.org/n?u=RePEc:zbw:pp1859:21&r=all |
| By: | Claudia Pacella |
| Abstract: | In this thesis I apply modern econometric techniques on macroeconomic time series. Forecasting is here developed along several dimensions in the three chapters. The chapters are in principle self-contained. However, a common element is represented by the business cycle analysis. In the first paper, which primarily deals with the problem of forecasting euro area inflation in the short and medium run, we also compute the country-specific responses of a common business cycle shock. Both chapters 2 and 3 deal predominately with business cycle issues from two different perspectives. The former chapter analyses the business cycle as a dichotomous non-observable variable and addresses the issue of evaluating the euro area business cycle dating formulated by the CEPR committee, while the latter chapter studies the entire distribution of GDP growth. |
| Keywords: | Inflation; Multi-country model; Forecasting; Bayesian estimation; Business fluctuations; Cycle; Factor model; Asymmetric least squares; Expectiles; Quantiles; Density forecasting |
| Date: | 2020–06–15 |
| URL: | http://d.repec.org/n?u=RePEc:ulb:ulbeco:2013/307579&r=all |
| By: | Nissilä, Wilma |
| Abstract: | This article surveys both earlier and recent research on recession forecasting with probit based time series models. Most studies use either a static probit model or its extensions in order toestimate the recession probabilities, while others use models based on a latent variable ap-proach to account for nonlinearities. Many studies find that the term spread (i.e, the difference between long-term and short-term yields) is a useful predictor for recessions, but some recent studies also find that the ability of spread to predict recessions in the Euro Area has diminished over the years. Confidence indicators and financial variables such as stock returns seem to provide additional predictive power over the term spread. More sophisticated models outper-form the basic static probit model in various studies. An empirical analysis made for Finland strengthens the findings of earlier studies. Consumer confidence is especially useful predictor of Finnish business cycle and the accuracy of the static single-predictor model can be improved by using multiple predictors and by allowing the dynamic extension. |
| Keywords: | business cycles,recession forecasting,probit models |
| Date: | 2020 |
| URL: | http://d.repec.org/n?u=RePEc:zbw:bofecr:72020&r=all |