BSP to shift to enhanced forecasting model by next year

The Bangko Sentral ng Pilipinas (BSP) aims to shift to an enhanced macroeconomic forecasting model by next year that will help in the formulation of monetary policy, officials said on Monday.

The BSP will adopt the Policy Analysis Model for the Philippines (PAMPh) for determining the path of monetary policy, which was made in partnership with the International Monetary Fund’s Institute for Capacity Development and the Japan International Cooperation Agency.

“The shift to conditional forecasts as a key input to monetary policy formulation is in line with best practices in inflation targeting central banks. Specifically, by representing the key economic gaps in the Philippine economy, the PAMPh allows a systematic, internally consistent, and economically coherent analysis of current and future economic conditions that will guide monetary policy formulation in the BSP,” BSP Department of Economic Research (DER) Economic and Financial Forecasting Group Deputy Director Dennis M. Bautista said at a briefing on Monday.

“Conditional forecasts will take the BSP in communicating current decisions and more importantly, the probable path of policy decisions helping to anchor inflation expectations,” he added.

The BSP will be shifting from its current multi-equation model (MEM) and single equation model (SEM).

The PAMPh generates a suggested policy path that Monetary Board members may use when making policy decisions. The previous models only forecast the inflation rate over a decided time frame based on a fixed policy rate.

“I think the key limitation is that the multi-equation model assumes a constant interest rate over the policy horizon,” BSP DER Director Dennis D. Lapid said. “So, this current model (PAMPh) ensures that when we give out a path for future inflation, it will have a corresponding path for not just GDP (gross domestic product), but also the policy interest rate along with it.”

The BSP will not disclose the policy paths forecasted by the PAMPh, and they will be for internal use for now, he added.

Mr. Bautista said the PAMPh uses 290 equations, while the MEM uses 24. The PAMPh also considers more sectors to see inflation drivers and other economic imbalances.

“The MEM has very good forecasting and performance but when you compare the two, the PAMPh considers more details of the economy. So, it can be used to address more, plus gives more explicit input on the monetary policy stance. It’s producing an entire path of the policy rate,” BSP Deputy Governor Francisco G. Dakila, Jr. said.

Mr. Bautista said the BSP will not be abandoning the old model but will use it along with the PAMPh.

“The MEM will then be one of the so-called satellite models. The PAMPh is quarterly because many of the variables are available only quarterly. But you need more detailed information on the development of inflation on a higher frequency basis, so that means you need a lot of so-called satellite or complementary models. The MEM will be one of them because it produces forecasts of inflation in the near term,” Mr. Dakila added.

University of the Philippines Los Baños Economics Senior Lecturer Enrico P. Villanueva said in a social media message that the main difference between the old and the new models could be better inflation forecasts.

“The past few months or year, their forecasted range of inflation, although a bit wide in range, normally captured the actual inflation figure,” he said.

“Due to the inherent difficulty of capturing and forecasting levels of economic variables, my suggestion is actually to temper the model’s output using expert judgment,” Mr. Villanueva added. “While macroeconomic modeling through graphical depiction is a doable and practical undertaking, its translation into mathematical or econometric models to be used in forecasting is not tractable. Econometric forecasting models are typically laden with assumption and oversimplification of complex real-life relationships among variables.” — AMCS