|
〕}} Nowcasting has recently become popular in economics. Standard measures used to assess the state of an economy, e.g., gross domestic product (GDP), are only determined after a long delay, and are even then subject to subsequent revisions. While weather forecasters know weather conditions today and only have to predict the weather tomorrow, economists have to forecast the present and even the recent past. Historically, nowcasting techniques have been based on simplified heuristic approaches. A recent paper by Giannone, Reichlin and Small (2008) has shown that the process of nowcasting can be formalized in a statistical model which produces predictions without the need for informal judgement. The model exploits information from a large quantity of data series at different frequencies and with different publication lags. The idea is that signals about the direction of change in GDP can be extracted from this large and heterogeneous set of information sources (e.g., jobless figures, industrial orders, the trade balance, etc.) before GDP itself is published. In nowcasting this data is used to compute sequences of current quarter GDP estimates in relation to the real time flow of data releases. Nowcasting models have been applied in many institutions, in particular Central Banks, and the technique is used routinely to monitor the state of the economy in real time. Selected academic research papers show how this technique has developed. Banbura, Giannone and Reichlin (2011) and Marta Banbura, Domenico Giannone, Michele Modugno & Lucrezia Reichlin (2013) provide surveys of the basic methods and more recent refinements. Nowcasting methods based on social media content (such as Twitter) have been developed to estimate hidden quantities such as the ('mood' of a population ) or the presence of a flu epidemic. A simple to implement regression-based approach to nowcasting involves mixed-data sampling or MIDAS regressions (see Andreou, Ghysels and Kourtellos (2011)〔Andreou, Elena & Eric Ghysels & Andros Kourtellos "Forecasting with Mixed-Frequency Data", Oxford Handbook of Economic Forecasting, Michael P. Clements and David F. Hendry (ed.) Chapter 8.〕). Mixed-data sampling (MIDAS) is an econometric regression or filtering method developed by Ghysels ''et al.'' There is now a substantial literature on MIDAS regressions and their applications, including Andreou et al. (2010),〔Andreou, Elena & Eric Ghysels & Andros Kourtellos "Regression Models with Mixed Sampling Frequencies", Journal of Econometrics, 158, 246-261.〕 and especially Andreou et al. (2013).〔Andreou, Elena & Eric Ghysels & Andros Kourtellos "Should macroeconomic forecasters use daily financial data and how?", Journal of Business and Economic Statistics 31, 240-251. 〕 The regression models can be viewed in some cases as substitutes for the Kalman filter when applied in the context of mixed frequency data. Bai, Ghysels and Wright (2013),〔Bai, Jennie, Eric Ghysels and Jonathan Wright "State Space Models and MIDAS Regressions" Econometric Reviews, 32, 779–813〕 examine the relationship between MIDAS regressions and Kalman filter state space models applied to mixed frequency data. In general, the latter involve a system of equations, whereas in contrast MIDAS regressions involve a (reduced form) single equation. As a consequence, MIDAS regressions might be less efficient, but also less prone to specification errors. In cases where the MIDAS regression is only an approximation, the approximation errors tend to be small. == References == 抄文引用元・出典: フリー百科事典『 ウィキペディア（Wikipedia）』 ■ウィキペディアで「Nowcasting (economics)」の詳細全文を読む スポンサード リンク
|