Inflation Rate Modelling Through a Hybrid Model of Seasonal Autoregressive Moving Average and Multilayer Perceptron Neural Network

This study examines the performance of seasonal autoregressive integrated moving average (SARIMA), multilayer perceptron neural networks (MLPNN), and hybrid SARIMA-MLPNN model(s) in modelling and forecasting inflation rate using the monthly consumer price index (CPI) data from 2010 to 2019 obtained from the South African Reserve Bank (SARB). The forecast errors in inflation rate forecasting are analyzed and compared. The study employed root mean squared error (RMSE) and mean absolute error (MAE) as performance measures. The results indicate that significant improvements in forecasting accuracy are obtained with the hybrid model (SARIMA-MLPNN) compared to the SARIMA and MLPNN. The MLPNN model outperformed the SARIMA model. However, the hybrid SARIMA-MLPNN model outperformed both the SARIMA and MLPNN in terms of forecasting accuracy/accuracy performance.

PERAMALAN INFLASI DI INDONESIA MENGGUNAKAN METODE AUTOREGRESSIVE MOVING AVERAGE (ARMA)

A good inflation rate for a country is an inflation rate that has a low and stable value so that able to realize fast and controlled economic growth. Forecasting can be one of the steps that can provide an overview of the value of inflation in Indonesia for the government or related agencies to formulate and maintain inflation stability in Indonesia. In this study, a forecasting analysis was carried out to determine the prediction of inflation in Indonesia in 2021 using the Autoregressive Moving Average (ARMA) method. From the results of the research that has been done, the best model to predict this case is using the ARMA model (3,0,0) because it produces the smallest AIC value of 0.2373 and the smallest RMSE of 7.81. From this model, the results of forecasting inflation rates for the months of May to December 2021 are also obtained with a range of 0.1% to 0.3%. The graphic pattern of the predicted results follows the actual data line pattern, which means that this model is good to use. Abstrak Tingkat inflasi yang baik bagi suatu negara adalah tingkat inflasi yang memiliki nilai yang rendah dan stabil, sehinga mampu mewujudkan pertumbuhan ekonomi yang cepat dan terkendali. Peramalan dapat menjadi salah satu langkah yang dapat memberikan gambaran nilai inflasi di Indonesia bagi pemerintah atau badan yang terkait untuk menyusun dan mempertahankan kestabilan inflasi di Indonesia. Dalam penelitian ini, dilakukan analisis peramalan untuk mengetahui prediksi angka inflasi di Indonesia tahun 2021 menggunakan metode Autoregresif Moving Average (ARMA). Dari hasil penelitian yang telah dilakukan, model terbaik untuk meramalkan kasus ini yaitu menggunakan model ARMA (3,0,0) karena menghasilkan nilai AIC paling kecil yaitu 0.2373 dan RMSE terkecil sebesar 7.81. Dari model tersebut juga didapatkan hasil peramalan angka inflasi untuk bulan Mei hingga Desember 2021 dengan kisaran 0.1% hingga 0.3%. Pola grafik dari hasil prediksi mengikuti pola garis data aktual yang berarti bahwa model ini baik untuk digunakan.

Inflation Expectations and Inflation: Nowcasting and Forecasting

Improving the mechanisms for forecasting inflation is an important part of economic science. National central banks, which monitor and manage the dynamics of the price level in the economy, use and develop these mechanisms in practice. Scientists and bank analysts have developed an impressive variety of ways to obtain estimates of inflation expectations of professional economists and ordinary citizens, as well as models for predicting future inflation values. In the last ten years, big data obtained from the Internet has been increasingly used for nowcasting inflation expectations and forecasting price dynamics. In this article, using the methods of correlation and regression analysis, it is demonstrated the validity of measuring the inflation expectations based on queries in Google Trends. In addition, these data turned out to be a fairly good predictor of the CPI level with a one-month lag. And combining the traditional CPI with a one-month lag and query statistics gave the lowest forecast error of all the model specifications considered. The resulting model is more flexible than the classical methods of forecasting inflation, including by taking into account the psychological aspects of economic behavior.

ТHE METHODOLOGY FOR INFLATION’ FORECASTING BY THE BANK OF RUSSIA IN THE MEDIUM TERM

The paper sets out to examine approaches to the forecasting of inflation by a macro market regulator. Various approaches to short-term inflation forecasting, inflation factors and their main channels of influence used by bank regulators in various countries are studied. The shortcomings of the used models for predicting inflation in the post-pandemic economy have been formulated. A comparative analysis of the use of various models has been conducted and solutions for building forecasting models in the medium term have been proposed. The approach has been tested for regional inflation forecasting; calculations of the indicators using VAR model, SARIMA, and dynamic method have been presented. It is proposed to use extended combined VAR models supplemented with exogenous factors for medium-term forecasting.

Forecasting Inflation in Latin American Countries Using a SARIMA-LSTM Combination

Inflation forecasting has been and continues to be an important issue for the world's economies. Governments, through their central banks, watch closely inflation indicators to make national decisions and policies. Controlling growth and contraction requires governments to keep a close eye on the rate of inflation. When planning strategic national investments, governments attempt to forecast inflation over longer periods of time. Getting the inflation forecast wrong, can result in significant economic hardships. However, even given its significance, there is limited new research that applies updated methodologies to forecast it, and even fewer studies in emerging economies where inflation may be drastically higher. This study proposes to forecast the inflation rate in emerging economies based on the commonly used Seasonal Autoregressive Integrated Moving Average (SARIMA) approach combined with Long Short Term Memory (LSTM). The results indicate that the proposed model based on the combination of SARIMA and LSTM, have a higher accuracy in inflation forecasts as measured by the Mean Square Error (MSE) of the proposed models over the SARIMA model and LSTM alone. The loss function used is Mean Squared Error (MSE), and the Model Confidence Set (MCS) is used to test the superiority of the models in the economies of Mexico, Colombia and Peru.

An Analysis on Forecasting Inflation Rate in the Philippines: A Recurrent Neural Network Method Approach

Inflation rate is the proceeding rise within the common level of costs of products and services in an economy over a certain span of time. In 2018, the Philippines has the highest inflation rate among the 10 South East Asian countries. The objective of this research is to forecast the inflation rate of the Philippines for the next five years (2019-2023). Also, the researchers compared the results obtained from the Multiple Linear Regression and Recurrent Neural Network (RNN) performed in MATLAB to determine which of these two models will be the better model in forecasting inflation rate. In this study, the researchers observed the behavior of the Inflation Rate(y) and its economic factors such as Import (x1), Export (x2), Money Supply (x3), Gross Domestic Product (x4), Gross National Product (x5), Expenditure (x6) and Exchange Rate (x7). Using Multiple Linear Regression, this study determined that the significant predictors are Money Supply (x3) and Expenditures (x6). By evaluating the forecast efficiency of the two methods, the researchers concluded that Multilayered Recurrent Neural Network outperforms Multiple Linear Regression in predicting inflation rate of the Philippines. This paper can be useful to the Philippine Government on their decisions about monetary policy making since forecasting the inflation rate has a huge importance and impact in conducting monetary policy.

Does the high-frequency data is helpful for forecasting Russian inflation?

Due to the fact that at the end of 2014 the Central Bank made the transition to a new monetary policy regime for Russia — the inflation targeting regime, the problem of forecasting inflation rates became more relevant than ever. In the new monetary policy regime, it is important for the Bank of Russia to estimate the future inflation rate as quickly as possible in order to take measures to return inflation to the target level. In addition, for effective monetary policy, the households must trust the actions of monetary authorities and they must be aware of the future dynamics of inflation. Thus, to manage inflationary expectations of economic agents, the Central Bank should actively use the information channel, publish accurate forecasts of consumer price growth. The aim of this work is to build a model for nowcasting, as well as short-term forecasting of the rate of Russian inflation using high-frequency data. Using this type of data in models for forecasting is very promising, since this approach allows to use more information about the dynamics of macroeconomic indicators. The paper shows that using MIDAS model with weekly frequency series (RUB/USD exchange rate, the interbank rate MIACR, oil prices) has more accurate forecast of monthly inflation compared to several basic models, which only use low-frequency data.