“Nowcasting” the Philippines (NowcastingPH) for better GDP growth forecasts

“Nowcasting” the Philippines (NowcastingPH) for better GDP growth forecasts

Economy

UnionBank Publication

UnionBank Publication

263 week ago — 4 min read

In a statement of Department of Finance (DoF) last December 2018, it said that “faulty” inflation forecasts made by analysts raised inflation expectations; thus, it contributes to the actual rise in price levels. Recently, the DoF also ranked the accuracy of analysts’ GDP forecasts. Out of the 20 analysts, UnionBank’s Economic Research Unit (ERU) was ranked 11th. (See table 1 below)

 

 The DoF used the “mean absolute percentage error” (MAPE) method, where the error is defined as actual or observed value minus the forecasted value, as percentage of the actual value. Thus, the smaller the MAPE, the better the forecast.

 

As a positive response, ERU has developed a GDP growth forecasting model called “MIDAS” or “Mixed-Data Sampling” regression analysis. Simply put, one can have a dependent variable that’s quarterly (like GDP, in this case), and a regressor or independent variables that's measured monthly, or even daily, frequency. (like inflation, liquidity, etc.). Typically, some “extra” values for the high-frequency variable/s will be available after the most recent value of the low-frequency dependent variable has been observed. These so-called “extra” observations can now be used and thus having the potential for what's usually termed as “nowcasting” in the forecasting literature. Think of nowcasting economic variables like how a meteorologist predicts the weather on a weekly- and daily-basis. 

 

From this, the ERU has called the forecasting model as ‘NowcastingPH’ or “Nowcasting Philippines”. NowcastingPH now has the potential to make GDP growth forecasts more accurate on a weekly- or monthly-basis, as long as new data is made available or is released by local statistics authorities.


As an initial result, NowcastingPH sees the first quarter of 2019 GDP growth to be at 6.4%, using the latest releases of the Philippine Statistics Authority. It is the hope that as extra high-frequency data or information is included that it will consequently improve the forecasting performance of the model. NowcastingPH will be released every time new economic data becomes available.

 

Article by Ruben Carlo Asuncion

 

Disclaimer: While this document is based on information obtained from sources we believe to be reliable, we do not make any representations as to its accuracy, completeness, correctness, timeliness or use for any particular purpose. Opinions and statements expressed here are those of their author(s) as of the date of this report and not of Union Bank of the Philippines (UBP). The opinions and statements provided in this document are subject to change without prior notice. Any recommendation contained in this document does not have regard to the reader’s particular investment objectives, financial situation and any other specific needs. This document is for informational purposes only and UBP is not soliciting any action based on it. Nothing here shall to any extent substitute for the independent investigations and the technical and business judgment of the reader. Your use of this document and any of its contents is at your own risk and UBP does not accept any liability for the results of any action or decision taken on the basis of or reliance on this document or any of its contents.

Comments

Posted by

UnionBank Publication

We are a team of professionals providing relevant content to startups, micro, small and medium enterprises.