
STOCHASTIC PREDICTION OF MONTHLY INFLATION RATES THROUGH KALMAN FILTERING
Author(s) -
G. A. DAWODU,
A. A. AKINTUNDE,
S. O. ARIYO
Publication year - 2019
Publication title -
journal of natural sciences, engineering and technology/journal of natural science, engineering and technology
Language(s) - English
Resource type - Journals
eISSN - 2315-7461
pISSN - 2277-0593
DOI - 10.51406/jnset.v16i2.1842
Subject(s) - univariate , kalman filter , inflation (cosmology) , econometrics , benchmark (surveying) , series (stratigraphy) , inflation rate , computer science , multivariate statistics , statistics , economics , mathematics , macroeconomics , interest rate , geodesy , geography , paleontology , physics , theoretical physics , biology
Inflation measure is an important indicator of the state of an economy and the desire to determine it ahead of “time” cannot be overemphasised. This paper presents a step-by-step algorithm to predict the would-be monthly inflation rate of the Nigerian economy, using Kalman Filtering Predictor (KFP). The ordinary structural model for a time series (structTS) is highlighted to “fairly” compete against our proposed KFP. The structTS is a powerful “competitor”, it is in recommended R package “stats” and used for fitting basic structural models to “univariate” time series. It is quite reliable and fast, and is used as a benchmark in some comparisons of filtering techniques, it is indeed the “predictor” to “beat”, yet our proposed KFP has more to “offer”. The pertinent statistics and pictorial representation of the results obtained, through both techniques, is highlighted for any “incorruptible” judge’s perusal. All of these are contained in the couple of illustrative examples that exhibit the steps involved in the proposed algorithm, using a hypothetical monthly inflation rate and the monthly inflation rates data (January, 2011 to June, 2014) of the Nigerian economy.