An optimal k of kth MA-ARIMA models under MA(q) models

DAWOUD I., KAÇIRANLAR S.

COMMUNICATIONS IN STATISTICS-SIMULATION AND COMPUTATION, vol.46, no.6, pp.4185-4198, 2017 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 46 Issue: 6
  • Publication Date: 2017
  • Doi Number: 10.1080/03610918.2015.1109657
  • Journal Name: COMMUNICATIONS IN STATISTICS-SIMULATION AND COMPUTATION
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Page Numbers: pp.4185-4198
  • Keywords: ARIMA models, Exponential weighted moving average, Forecasting accuracy, Simple moving average, Weighted moving average
  • Çukurova University Affiliated: Yes

Abstract

In this article, we discuss finding the optimal k of (i) kth simple moving average, (ii) kth weighted moving average, and (iii) kth exponential weighted moving average based on simulated MA(q) model. We run a simulation using the three above examining methods under specific conditions. The main finding is that, 5th Exponential Weighted Moving Average (5-th EWMA) Autoregressive Integrated Moving Average (ARIMA) model is the best forecasting model among others, which means the optimal k = 5. For Turkish Telecommunications (TTKOM), stock market real data reveals the similar results of the simulation study.